_ U — Ls 5 i ee Sore Lien, Vy" - oe nse ar \ oJ) be 4S fe7F sa 25

& w = wn = w q

SStYWYEIT_ LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS SJ ®

= = w ae . 22) = n” i

79.5 Ne 3 5 \a i £%2 z §

lp 5 a EN S 2 a NY 2 Vip a SY

aid = AY SS z Es AS eg eee = =

nO Ba Co ee 5 2 =

NSTITUTION NOHNLILSNI_NVINOSHLIWS Saluvaa!y LIBRARIES SMITHSONIAN INS ;

” = = :

ts a ul = gy, 2k

oc oc ti om Bi a .

-¢ c . = <x 4 2 j

ba = \ oO esr lead a par y

= ih epee ene rs) = o “Ms

oe ¥3 aah = aul =

sJTYVYEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS Saly

= am ra = IK: z 3

= O ee © =~ Oo :

ei rag ¢@ ri w = 3

0 Ff a = ae =

7a GHZ a aaa oe) cs

ata = “Gy — o a Ke

% ae eee z is Z

NSTITUTION NOILNLILSNI NVINOSHLIWS Sd1uYvugiy INS

’ ” = w Zz ” =

« Fz <x = 2 = =

SS oN mail f — 4% z= _ —

We F é 2 é = z

2 I = = 2 ‘3

> = > : = oe =

§> 7) 2 ” Ai ae n

J1uvysi SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS Say

S wn > ” > Let ”

” , Pa?) uf es . WW 5

~{ e. ttf So a ee 4 a ‘

= :‘ Bip 5 < 5 =

= " tps C o <x

4 ae GG aa i 4 -

S a GF os . 5 o Nea

im wl ra =! a somes

STITUTION NOILALILSNI NVINOSHLINS S3IlyuYVvygiy LIBRARIES SMITHSONIAN INSTIT

z we ee = 7 z :

Oo Tas ae : at @

_ ow we oo — @

= 2 Nag 5 » 5 G2B >»

E aS = . Se di a

fos eed NN 98 Pia ra ‘ Of 7-4 od

IYVUYGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS S31

F ee ‘~ wo za r w” = 2]

Vy = : S = = < = y

£43 Ne 3 Z = a 2

Las fy N ” ” ) sel n ie

ek ET Sg WN 2) ea ee te Oo

fr - s \ \ > a ae: ia =

= ‘s > = >” = >

” ap 2 ” ie z ” an .

STITUTION NOILNLILSNI NVINOSHLIWS SJIYVYEIT LIBRARIES SMITHSONIAN INSTIT

ap) > _ io) = w” y S :

oc - o at. ~ re a

< Cc << eo < ; a

oc “4 ac ear o a sal

a O S i 3 ee fs)

cs & xo = ood Zz

1'YVYdIT LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLIWS S3tu

» z i z BY | z

n = A y ra xD e &

> b 33 a be “el E

2 - a i c =

TITUTION NOILNLILSNI NWINOSHIINS SA1YWY EIT LIBRARIES SMITHSONIAN INSTIT

== QS Ss¥ Ww SSB, _— 4, SS . ms oa EE eae § — 1 34 = <2. Gare

m SS Was m LI n° m tins ” SS Was

1ES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3SIYVYUSIT LIBRA

2 a eer Z ae ,

= = = > = ty .

: : : 3 : > x

a) ” ‘ Me So n Up pa 8

Oo are oO Bi 2 O A £Foy ,: YQ

= =, = = a \

ol = ~~: > * => >’ 2)

2 ” art Ps ” z=

NVINOSHLIWS S3J!1YVYEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNJ

w“Y

A ul 2 2 2 e

pa — ao ee oc

= _< #3 <x a SN x

y 4 ec = aad se oO

Wwash~ Oo = Oo ae Oo =

Beer 23: oe a z el aa = a"

ARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3INYUgI] LIBRA

bat! 2 J z ce , z Sa

Sco = OO = = eS) a

£ > wa > >) ke BY

M s = ‘= 2 a ee

Non oc. ” ue 2) a

<a. o z : ie 0

DILALILSNINVINOSHLINS S3INVUEIT_ LIBRARIES SMITHSONIAN INSTITUTION, NO!N

a = @ = = aX =

= 4 ae 4 = WA: Be

Ww GD g WY) w~ Y”) ~

= Og Se oO oe oO

= = = z = z

BRARIES SMITHSONIAN INSTITUTION NOILMLILSNI NVINOSHLIWS Salyvuait_L!BR/

e a = n = A

ae 5 aie = 8 = aa ae

Par a x = ae i

ie Cc a c ac =

AC: SF = mM =f pats

Nowoe% © : Oo _ O = S

q = 5 Ps a | ros ~ BS!)

DILALIZSNI NVINOSHLINS S3IYVYUSIT LIBRARIES SMITHSONIAN” INSTITUTION

ia > rats is ee Se =

a ES Pi = i i

ow = ow — Vf ow ae

: = : - Gly = :

> a > a LL: =

= = Z = se fd) ve

Nant om SS ob = my? | ee 2

; w : a W = wn Sap

BRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS SaluvudiT_LIBR?

= ae z < = j = x

J + =

=: S =. S ~ SS

WK S ” LN 4 oe a - en

. = 3 S = Zy} = fe,

oon 2 PF 2 a 2 ® wot.

OILNLILSNI_ NWINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAN INSTITUTION |

a ul z Me : Z ul

s re = =: = a

5 ar ae fo <

= Xs _< = < Tikes oc

BRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NWINOSHLIWS saiuvadi LIBR?

< z +: rs 5 . = a

= Ie = . i= Fe)

= ne = A a >

E : bad E cae i 5)

E a = ta = —

tii ” w”~

a eee i z - ee: o

NOLLNLILSNI_ NVINOSHLINS $3 luVUGIT_LIBRARI ES SMITHSONIAN _ INSTITUTION , NOILN

i xf KWON OS fee ohare sh a: ee” i =

a7 7.65

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

VOLUME 65

THE TRUSTEES OF THE DIE TRUSTEES VAN DIE

SOUTH AFRICAN MUSEUM SUID-AFRIKAANSE MUSEUM

CAPE TOWN KAAPSTAD

1974

# PRINTED IN THE REPUBLIC OF SOUTH AFRICA BY

THE RUSTICA PRESS (PTY.) LTD., WYNBERG, CAPE

436

- ANNALS OF THE

ir SOUTH AFRICAN MUSEUM

VOLUME 65

NEW GENERIC NAMES

PROPOSED IN THIS VOLUME

Cunicus Griffiths, 1974

Dikwa Griffiths, 1974

Mesembriportax Gentry, 1974

LIST OF CONTENTS

BOUILLON, J.

See

MILLARD, N. A. H. & BOUILLON, J.

CONNELL, A. D. & GRINDLEY, J. R.

Two new species of Acartia (Copepoda, Calanoida) from South African estuaries

(published July 1974) he :

FOuRIE, S.

The cranial morphology of Thrinaxodon liorhinus Seeley (published October 1974)

GENTRY, A. W.

A new genus and species of Pliocene bosephaline (Bovidae, Mammalia) from

South Africa (published July 1974) :

GRIFFITHS, C. L.

The Amphipoda of southern Africa. Part 4. The Gammaridea and Caprellidea of the

Cape Province east of Cape Agulhas (published August 1974) :

GRINDLEY, J. R.

See

CONNELL, A. D. & GRINDLEY, J. R.

Harris, J. M.

Orientation and variability in the ossicones of African Sivatheriinae (Mammalia:

Giraffidae) (published July 1974) ..

KENSLEY, B. F.

Aspects of the biology and oa of the genus T A Latreille oe mas

October 1974) : 3 ate

KENSLEY, B. F.

The status of the Plio-Pleistocene Panopea in southern Africa (Mollusca, Bivalvia,

Hiatellidae) (published July 1974)

KLEIN, R. G.

On the taxonomic status, distribution and ecology of the blue antelope, Hippo-

tragus leucophaeus (Pallas, 1766) (published July 1974)

MILLARD, N. A. H. & BOUILLON, J.

A collection of hydroids from Mocgambique, East Africa (published March 1974)

RAU, R. E.

Revised list of the preserved material of the extinct Cape Colony quagga, Equus

quagga quagga (Gmelin) (published May 1974) .. : ye

TANKARD, A. J.

Petrology and origin of the phosphorite and aluminium phosphate rock of the

Langebaanweg-—Saldanha area, south-western Cape Province (published July

1974) ee:

PAGE

337

145

251

189

401

199

247,

it r ine

rca | : 4

AS Voda

; 2

Pig

fee i } q i 7 i ;

aii 4 ‘ a 2 ft eee 4

‘OF THE SOUTH AFRICAN

MES

CAPE TOWN |

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 Band

March 1974 Maart

Part 1 Deel

A COLLECTION OF HYDROIDS

FROM MOCAMBIQUE, EAST AFRICA

N. A. H. MILLARD

J. BOUILLON

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 2, 3, 5, 7-8), 31-2, 5, t.=p.1.), 511-2, 5, 7-9),

6(1, t.-p.i.), 7-3), 8, 9-2), 10(1),

11(1-2, 5, 7, t.—p.i.), 24(2), 27, 31(1-3), 33

Price of this part/Prys van hierdie deel

R6,00

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 37 2

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA

N. A. H. MILLARD

South African Museum, Cape Town

J. BOUILLON

Université Libre de Bruxelles

(With 9 figures)

[MS accepted 30 May 1973]

CONTENTS

PAGE

Introduction : : ; 5 : 1

List of materials and localities. ; : 2

Systematic section . : : : : 10

Summary ; : : ; : : 38

Acknowledgements : : : 38

References 2 ; A : : f 38

INTRODUCTION

This paper is the second part of an account of the collection made by the

second author and his colleagues during expeditions to the Seychelles in 1960

and to the east coast of Africa in 1969. The hydroids of the Seychelles, which

comprised the first part, are described by Millard & Bouillon (1973).

Most of the collection comes from the southern part of Mocambique:

from Inhaca Island and Ilha dos Portugueses, which lie to the east of Lourencgo

Marques in Delagoa Bay, but there is also material from Santa Carolina, a

small island further north between the larger Ilha do Bazaruto and the mainland.

The area thus lies on the border of the tropics and is in general under the

influence of the warm southward-flowing Mocambique current, although the

temperature of the coastal waters is somewhat less due to the presence of an

inshore northward-flowing counter-current. Macnae & Kalk (1958) give a

temperature of approximately 26°C for the Mocambique current and 20°C

for the eastern shores of Inhaca Island, whereas the water in Delagoa Bay itself

tends to be warmer due to the heating of the sun during the day.

The position of the collecting stations is listed as follows:

Inhaca, Station. On the west coast of Inhaca Island. 26°03’S/32°54’E.

Inhaca, Barreira Vermelha. On the west coast of Inhaca Island.

26°02'S/32°54’E.

Ann. S. Afr. Mus. 65 (1), 1974: 1-40, 9 figs.

2 ANNALS OF THE SOUTH AFRICAN MUSEUM

Ilha dos Portugueses. An island to the north-west of Inhaca Island, in

Delagoa Bay. 25°58'S/32°55 E.

Inhaca Est.

Inhaca, Saco. A bay on the south coast of Inhaca Island. 26°03’S/32°56’E.

Cabo da Inhaca. The most north-easterly point of the island.

25° 58.5) 3275978:

Inhaca Est (dragage, 10-15 m).

Inhaca, Ponta Torres. The most south-easterly point of the island.

26-05/S/32 ome.

Santa Carolina. An island further north on the coast of Mocgambique.

23T Si 5520) 1s

An account of the marine ecology of Inhaca Island will be found in Macnae

& Kalk (1958), while some of the hydroids from the area were described by

Millard (1958, 1959).

Since many of the species from southern Africa have already been described

or are too well known to warrant description, only certain ones are described

in detail and illustrated. All are recorded in the list which follows. As in the

previous paper, unidentifiable material is not recorded.

The expedition was financed by the Université Libre de Bruxelles, the Musée

Royal de lAfrique Centrale (Tervuren), the Centre Belge d’Océanographie

and the Belgian Ministry of Education and Culture.

The collection is the property of the Musée Royal de I’Afrique Centrale,

where the type of the single new species has been deposited.

LIST OF MATERIAL AND LOCALITIES

*Species marked with an asterisk are described or discussed in the systematic

section.

+Species marked with a dagger are new records for Africa south of 20°S. Lat.

Family Myriothelidae

Myriothela sp.

Santa Carolina: a single infertile specimen 0,8 mm in total height.

Family Tubulariidae

*Ectopleura bethae (Warren, 1908)

Inhaca: a rich, fertile colony.

Zyzzyzus solitarius (Warren, 1906)

Cabo da Inhaca: several hydranths, one fertile, growing in a sponge.

Santa Carolina: one fertile hydranth in sponge.

Family Cladocorynidae

* Cladocoryne floccosa Rotch, 1871

Ihla dos Portugueses: a rich fertile colony on Dynamena crisioides.

Inhaca, Saco: an infertile colony on Lytocarpus philippinus.

Inhaca, Ponta Torres: a rich colony with a few fertile hydranths.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA

Family Halocordylidae

Halocordyle disticha (Goldfuss, 1820)

Inhaca, Ponta Torres: stems reaching 56 mm, some with young

gonophores.

Santa Carolina: fertile stems reaching 35 mm.

Family Corynidae

*Coryne ?pusilla Gartner, 1774.

Inhaca Est: an infertile colony on weed.

Inhaca, Ponta Torres: infertile colonies.

* Sphaerocoryne bedoti Pictet, 1893

Santa Carolina: a single infertile hydranth detached from substratum.

Family Cladonemidae

+ ?Cladonema sp.

Inhaca: three young, infertile hydranths reaching 0,8 mm.

Family Solanderiidae

+ Solanderia minima (Hickson, 1903)

Santa Carolina: an infertile colony 70 mm in height and 70 mm in

spread.

Family Zancleidae

+*Zanclea sp.

Inhaca, coral reef west of Station: an infertile colony commensal on a

coral (colony 1).

Inhaca, coral reef west of Station: a fertile colony from coral reef,

growing with a polyzoan (colony 2).

Inhaca, Barreira Vermelha: two colonies, one fertile, on polyzoan

(colonies 3 and 4).

Family Clavidae

Corydendrium parasiticum (“innaeus, 1767)

Inhaca, Station: an infertile colony reaching 46 mm.

Inhaca Est: a young colony reaching 8 mm.

Inhaca, Saco: a rich, but infertile, colony reaching 56 mm.

Inhaca, Ponta Torres: rich, but infertile, colonies reaching 59 mm.

+* Rhizogeton nudum Broch, 1909

Inhaca, Station: a fertile colony on weed.

Ihla dos Portugueses: a few infertile hydranths.

Santa Carolina: a few infertile hydranths.

+* Turritopsis nutricula (McCrady, 1856)

Inhaca, Station: a young, infertile colony.

Inhaca Est (dragage): a few young hydranths.

Santa Carolina: several small colonies, one with young medusae.

Family Cytaeidae

Cytaeis nassa (Millard, 1959)

Ilha dos Portugueses: fertile colonies on shells of Nassa fenestrata.

4 ANNALS OF THE SOUTH AFRICAN MUSEUM

Family Eudendriidae

* Fudendrium capillare Alder, 1856

Inhaca, Barreira Vermelha: one infertile colony.

Cabo da Inhaca: one infertile colony.

Inhaca, Ponta Torres: three fertile colonies.

Santa Carolina: one infertile colony.

+* Eudendrium motzkossowskae Picard, 1951

Inhaca, Saco: one infertile colony.

Inhaca, Ponta Torres: three colonies, two male and one infertile.

Santa Carolina: one infertile colony.

* Eudendrium ramosum (Linnaeus, 1758)

Inhaca Est: four colonies, three of them fertile.

Inhaca, Saco: one infertile colony.

Cabo da Inhaca: one infertile colony.

Inhaca, Ponta Torres: eight colonies, five of them fertile.

Family Hydractiniidae

* Hydractinia diogenes Millard, 1959

Inhaca, Station: a fertile colony on hermit shell.

Inhaca, Ponta Torres: a fertile colony on empty shell.

*Podocoryne ?carnea M. Sars, 1846

Inhaca Est (dragage): a female colony on weed.

*Incertae sedis (1)

Inhaca Est (dragage): a few minute hydranths on polyzoan.

*Incertae sedis (2)

Inhaca, Barreira Vermelha: an infertile colony.

t*Incertae sedis (3) Lineolaria sp.

Inhaca, Barreira Vermelha: an infertile colony.

Family Aequoreidae

Aequorea africana Millard, 1966

Ilha dos Portugueses: an infertile colony with a few contracted

hydranths.

Family Calicellidae

+ Egmundella amirantensis Millard & Bouillon, 1973

Inhaca, Station: two infertile colonies.

[hla dos Portugueses: infertile colonies on Dynamena crisioides.

Santa Carolina: an infertile colony on Dynamena crisioides.

Family Haleciidae

Halecium halecinum (Linnaeus, 1758)

Inhaca Est (dragage): a small female colony reaching 4 mm, on weed.

Halecium inhacae Millard, 1958

Inhaca, Station: a fertile colony reaching 4 mm.

Inhaca Est: fertile colonies on weed, reaching 3 mm.

Inhaca, Ponta Torres: rich fertile colonies on weed, reaching 3,5 mm.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA

Halecium lankesteri (Browne, 1890)

Santa Carolina: a small male colony reaching 2,6 mm, on worm tubes

and other hydroids, and a rich female colony reaching 1,5 mm, on

weed.

* Halecium tenellum Hincks, 1861

Ilha dos Portugueses: an infertile colony reaching 3 mm.

Cabo da Inhaca: fertile colonies reaching 5 mm.

Inhaca Est (dragage): an infertile colony reaching 5 mm.

Inhaca, Ponta Torres: a rich fertile colony reaching 3 mm.

+* Hydrodendron gardineri (Jarvis, 1922)

Cabo da Inhaca: several infertile colonies on the ascidian Pyura

and on the mussel Perna perna.

* Hydrodendron sympodiformis n. sp.

Inhaca, Ponta Torres: a very rich fertile colony on weed.

Family Campanulariidae

Campanularia crenata (Hartlaub, 1901)

Inhaca, Station: an infertile colony on weed.

Inhaca, Barreira Vermelha: a fertile colony on weed.

Inhaca, Ponta Torres: a rich fertile colony on weed.

* Campanularia delicata (Trebilcock, 1928)

Cabo da Inhaca: fertile colonies on weed.

Inhaca Est (dragage): an infertile colony on weed.

Campanularia integra MacGillivray, 1842

Inhaca Est: rich fertile colonies on weed.

Inhaca Est (dragage): a rich fertile colony on weed.

Inhaca, Ponta Torres: rich fertile colonies on weed.

Campanularia laminacarpa Millard, 1966

Cabo da Inhaca: an infertile colony epizootic on Thyroscyphus aequalis.

Campanularia morgansi Millard, 1957

Cabo da Inhaca: fertile colonies on Pyura and Perna perna.

Clytia gravieri (Billard, 1904)

Inhaca, Station: two colonies, one fertile, with branched and stolonial

stems.

Inhaca, Barreira Vermelha: a fertile colony with stolonial and slightly~

branched stems, on weed.

Thla dos Portugueses: an infertile stolonial colony.

Cabo da Inhaca: slightly branched stems with one gonotheca.

Inhaca, Ponta Torres: slightly branched infertile stems reaching 14 mm.

Santa Carolina: several stolonial colonies on weed, some fertile.

Clytia hemisphaerica (Linnaeus, 1767)

Inhaca, Station: several colonies, some fertile, with gonothecae of the

Johnstoni type.

Inhaca, Barreira Vermelha: fertile colonies on weed with gonothecae

of the johnstoni type.

6 ANNALS OF THE SOUTH AFRICAN MUSEUM

Ihla dos Portugueses: fertile colonies on weed with gonothecae of the

Johnstoni type.

Cabo da Inhaca: two infertile colonies.

Inhaca Est (dragage): infertile colonies on weed.

Inhaca, Ponta Torres: several infertile colonies.

Santa Carolina: several colonies on weed, one fertile, with gonothecae

of the johnstoni type.

+Clytia latitheca Miilard & Bouillon, 1973

Cabo da Inhaca: a small infertile colony reaching 6 mm.

Clytia paulensis (Vanh6ffen, 1910)

Inhaca, Saco: an infertile colony epizootic on Lytocarpus philippinus.

*Clytia sp.

Inhaca Est (dragage): a small infertile colony on weed.

Obelia bicuspidata Clarke, 1875

Inhaca, Station: infertile stems reaching 4 mm, on weed.

Ihla dos Portugueses: infertile stems reaching 5 mm.

Obelia dichotoma (Linnaeus, 1758)

[hla dos Portuguese: rich fertile colonies reaching 9 mm, on weed and

epizootic on Dynamena crisioides.

Cabo da Inhaca: infertile colonies reaching 8 mm.

Inhaca, Ponta Torres: rich fertile colonies reaching 7 mm, on weed.

Santa Carolina: an infertile colony reaching 7 mm.

Obelia geniculata (Linnaeus, 1758)

Inhaca, Station: a rich infertile colony on weed.

Inhaca Est: rich fertile colonies reaching 6 mm, on weed.

Cabo da Inhaca: a fertile colony on a gastropod shell.

Inhaca, Ponta Torres: fertile colonies reaching 5 mm, on weed.

Family Lafoeidae

Hebella furax Millard, 1957

Inhaca, Barreira Vermelha: an infertile colony epizootic on Halopteris

polymorpha.

Ihla dos Portugueses: an infertile colony epizootic on Halopteris

polymorpha.

Hebella scandens (Bale, 1888)

[hla dos Portugueses: infertile colonies epizootic on Dynamena

crisioides and D. cornicina.

Cabo da Inhaca: an infertile colony epizootic on Dynamena quadri-

dentata.

Inhaca Est (dragage): a rich colony epizootic on Sertularia linealis,

with a few gonothecae.

Inhaca, Ponta Torres: an infertile colony epizootic on Dynamena

crisioides.

Santa Carolina: infertile colonies epizootic on Dynamena crisioides.

—~

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA

* Scandia mutabilis (Ritchie, 1907)

Inhaca Est: an infertile colony epizootic on Eudendrium ramosum.

Cabo da Inhaca: infertile colonies epizootic on various hydroids.

Inhaca, Ponta Torres: infertile colonies epizootic on Halocordyle

disticha, Eudendrium ramosum and Idiellana pristis.

Family Syntheciidae

Hincksella corrugata Millard, 1958

Inhaca Est (dragage): one infertile stem of 3 mm.

Santa Carolina: an infertile colony reaching 10 mm.

Synthecium ?elegans Allman, 1872

Santa Carolina: five infertile stems reaching 15 mm.

Family Sertulariidae

* Abietinaria laevimarginata (Ritchie, 1907)

Ihla dos Portugueses: fertile colony on weed.

Inhaca Est (dragage): infertile colonies on weed.

Amphisbetia minima (D’Arcy Thompson, 1879)

Inhaca Est: rich fertile colonies reaching 3 mm, on weed.

Inhaca Est (dragage): an infertile colony reaching 2,5 mm, on weed.

Inhaca, Ponta Torres: rich fertile colonies reaching 3 mm, on weed.

+* Diphasia digitalis (Busk, 1852)

Inhaca Est: a branched stem of 7 mm and three smaller unbranched

ones, all infertile.

Inhaca, Ponta Torres: infertile branched and unbranched stems

reaching 44 mm.

* Diphasia tetraglochina Billard, 1907

Inhaca Est: an infertile colony reaching 2,5 mm, on weed.

Cabo da Inhaca: an infertile colony reaching 7 mm.

Inhaca Est (dragage): an infertile colony reaching 3 mm, on weed.

Dynamena cornicina McCrady, 1858

Ihla dos Portugueses: unbranched fertile stems reaching 4 mm.

Inhaca Est: two unbranched infertile stems reaching 9 mm.

Santa Carolina: several branched infertile stems reaching 60 mm,

and two rich fertile colonies with unbranched stems reaching 9 mm..

* Dynamena crisioides Lamouroux, 1824

Ihla dos Portugueses: many tall infertile stems of var. gigantea reaching

159 mm.

Inhaca, Saco: two infertile stems of normal variety reaching 13 mm.

Inkaca, Ponta Torres: infertile stems of normal variety reaching 20 mm.

Santa Carolina: many stems, reaching 160 mm, some fertile, including

normal variety, var. gigantea and var. alternata.

Dynamena obliqua Lamouroux, 1816

Inhaca Est (dragage): an infertile colony reaching !1 mm, on weed.

Inhaca, Ponta Torres: infertile stems reaching 4 mm, on weed.

8 ANNALS OF THE SOUTH AFRICAN MUSEUM

Dynamena quadridentata (Ellis & Solander, 1786)

[hla dos Portugueses: a fertile female colony reaching 6 mm, on weed.

Inhaca, Saco: infertile stems reaching 5 mm.

Cabo da Inhaca: infertile stems reaching 7 mm, on Pyura, Perna

perna and weed.

Inhaca Est (dragage): infertile stems reaching 5 mm, on weed.

Inhaca, Ponta Torres: infertile stems reaching 4 mm, on weed.

Santa Carolina: infertile stems reaching 3 mm, on weed.

Idiellana pristis (Lamouroux, 1816)

Inhaca, Station: several young colonies reaching 21 mm, on a bivalve.

Inhaca, Barreira Vermelha: two infertile stems reaching 19 mm.

Inhaca, Ponta Torres: infertile stems reaching 52 mm.

Santa Carolina: two infertile stems reaching 45 mm.

Sertularella mediterranea asymmetrica Millard, 1958

Inhaca Est: a fertile colony reaching 6 mm.

Inhaca, Ponta Torres: a fertile colony reaching 7 mm, on weed.

Sertularia distans (Lamouroux, 1816)

Inhaca, Station: unbranched stems reaching 12 mm, some fertile,

on weed and a shell.

Inhaca, Ponta Torres: unbranched, infertile stems reaching 4 mm,

on weed.

Santa Carolina: unbranched, infertile stems reaching 4 mm, on weed.

Sertularia ligulata Thornely, 1904

Inhaca, Station: infertile stems reaching 27 mm.

Inhaca Est (dragage): infertile stems reaching 8 mm, on weed.

Santa Carolina: infertile stems reaching 10 mm.

* Sertularia longa (Millard, 1958)

Ihla dos Portugueses: fertile colonies reaching 3 mm, on weed.

Inhaca Est (dragage): infertile colonies on weed.

* Sertularia marginata (Kirchenpauer, 1864)

Inhaca, Ponta Torres: infertile stems reaching 16 mm.

Sertularia turbinata (Lamouroux, 1816)

Inhaca, Station: an infertile colony reaching 6 mm.

Thyroscyphus aequalis Warren, 1908

Cabo da Inhaca: several infertile stems reaching 39 mm.

Thyroscyphus fruticosus (Esper, 1793)

Inhaca, Station: fertile fragments reaching 50 mm.

Family Plumulariidae

Antennella secundaria (Gmelin, 1791)

Inhaca Est: infertile stems reaching 5 mm.

Inhaca, Saco: infertile stems reaching 12 mm.

Inhaca, Ponta Torres: stems reaching 10 mm, some fertile.

Santa Carolina: stems reaching 8 mm, some fertile.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 9

+ Dentitheca bidentata (Jaderholm, 1920)

Inhaca, Barreira Vermelha: infertile stems reaching 11 mm.

Halopteris glutinosa (Lamouroux, 1816)

Inhaca Est: many stems, some fertile, reaching 17 mm.

Inhaca, Ponta Torres: stems reaching 18 mm, some fertile.

Halopteris polymorpha (Billard, 1913)

Inhaca Station: fertile stems reaching 11 mm.

Inhaca, Barreira Vermelha: an infertile stem of 9 mm.

[hla dos Portugueses: stems reaching 16 mm, some fertile.

Inhaca, Ponta Torres: an infertile stem of 16 mm.

Santa Carolina: many infertile stems reaching 22 mm.

Kirchenpaueria irregularis (Millard, 1958)

Santa Carolina: two infertile colonies with fascicled branched stems

reaching 13 mm.

+ Monostaechas quadridens (McCrady, 1857)

Cabo da Inhaca: stems reaching 13 mm, some fertile, on Pyura,

Perna perna and a sponge.

Inhaca, Ponta Torres: infertile stems reaching 8 mm.

Santa Carolina: many infertile stems reaching 13 mm.

Oswaldella nova (Jarvis, 1922)

Santa Carolina: infertile colonies epizootic on Halopteris polymorpha.

Plumularia filicaulis Kirchenpauer, 1876

Inhaca Est: fertile stems reaching 5 mm, on weed.

Inhaca Est (dragage): an infertile colony reaching 2,5 mm, on

weed.

Inhaca, Ponta Torres: fertile colonies reaching 5 mm, on weed.

+*Plumularia obliqua (Johnston, 1847)

Inhaca Est: an infertile colony reaching 4 mm, on weed.

Inhaca Est (dragage): an infertile colony reaching 5 mm, on

weed.

+Plumularia pennycuikae Millard & Bouillon, 1973

Santa Carolina: two infertile colonies reaching 13 mm.

Plumularia setacea (Linnaeus, 1758)

Cabo da Inhaca: abundant fertile stems reaching 22 mm, on Perna

perna, sponges and other substrata.

Santa Carolina: fertile stems reaching 20 mm.

+Plumularia strictocarpa Pictet, 1893

Santa Carolina: several fertile colonies reaching 10 mm.

Plumularia warreni Stechow, 1919

Inhaca, Ponta Torres: a fertile colony reaching 5 mm, on weed.

Santa Carolina: fertile colonies reaching 15 mm.

Plumularia wasini Jarvis, 1922

Santa Carolina: an infertile colony reaching 13 mm.

10 ANNALS OF THE SOUTH AFRICAN MUSEUM

Pycnotheca mirabilis (Allman, 1883)

Inhaca, Barreira Vermelha: infertile stems reaching 13 mm.

Inhaca Est (dragage): two infertile stems reaching 6 mm.

Inhaca, Ponta Torres: a fertile colony reaching 14 mm.

+* 4glaophenia cupressina Lamouroux, 1816

Santa Carolina: several fertile colonies reaching 260 mm.

Gymnangium gracilicaule gracilicaule (Jaderholm, 1903)

Inhaca, Station: an infertile colony reaching 94 mm.

*Gymnangium gracilicaule lignosum (Millard, 1968)

Santa Carolina: several infertile stems reaching 35 mm.

Lytocarpus philippinus (Kirchenpauer, 1872)

Inhaca, Station: infertile colonies reaching 41 mm, on weed.

[hla dos Portugueses: one infertile stem of 104 mm.

Inhaca, Saco: luxurious infertile stems reaching 147 mm.

Inhaca, Ponta Torres: luxurious infertile stems reaching 112 mm.

Santa Carolina: infertile fragments and young colonies.

+Thecocarpus delicatulus (Busk, 1852)

Santa Carolina: infertile stems reaching 30 mm.

Family Proboscidactylidae

+* Proboscidactyla (Lar) sp.

Inhaca, Barreira Vermelha: five colonies, one fertile.

SYSTEMATIC SECTION

Family Tubulariidae

Ectopleura bethae (Warren, 1908)

Tubularia betheris Warren, 1908: 280; pl. 45, figs 10, 11; pl. 46, fig. 12.

Description

A colony of many individuals reaching a maximum height of 30 mm.

Perisarc smooth or irregularly annulated in some areas, stiff to about 0,5 mm

below hydranth body, then becoming thin and membranous and terminating

on the basal dilation. Mature hydranth about 1,5 mm in height to tip of hypo-

stome, with up to 15 aboral tentacles and 17 oral tentacles. Aboral tentacles

reaching 1,8 mm, oral tentacles reaching 0,5 mm.

In the young hydranth the oral tentacles are definitely capitate with a

terminal battery of nematocysts. In the mature one the capituli are no longer

distinct but there is still a concentration of nematocysts on the distal end. The

smallest hydranth present is 0,42 mm in height and has nine filiform aboral

tentacles 0,24 mm long, and nine capitate oral tentacles 0,07 mm long.

Medusa-buds borne on branching blastostyles just above the aboral tenta-

cles, the largest 0,24 mm in height and 0,23 mm in diameter, with four capitate

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 1]

marginal tentacles 0,06 mm in length and eight exumbrellar bands of nemato-

cysts. In all of them the hypostome is large and fills the umbrellar cavity, and in

a few it protrudes through the mouth aperture. It has a circle of nematocysts

around the rim.

Remarks

This species resembles Ectopleura dumortierii (Van Beneden) in the presence

of four medusa tentacles, but differs from it in the smaller number of hydranth

tentacles and in the capitate tentacles of the young hydranth.

Family Cladocorynidae

Cladocoryne floccosa Rotch, 1871

Fig: DLE

Cladocoryne floccosa: Warren, 1908: 284. Behner, 1914: 419, figs 19-23. Philbert, 1936: 1,

figs 1-8. Weill, 1937: 1, figs 1-4. Vervoort, 1941: 190. Brinckmann-Voss, 1970: 69,

figs 80-82.

Description

Stems unbranched or rarely with one lateral branch, reaching a maximum

height of 5,0 mm. Perisarc smooth, or annulated in basal region only. Hydranths

with 4-7 oral tentacles and up to 18 branched aboral tentacles.

Male gonophores borne on hydranths amongst the aboral tentacles, these

hydranths showing no signs of regression to gonozooids as described by Behner.

Gonophores spherical, reaching 0,34 mm in diameter.

Nematocysts of three types:

(i) Macrobasic euryteles, found on body of hydranth only (Fig. | E).

Capsule bean-shaped, 28,8-31,2 x 11,7-15,0 u. Butt in two sections,

the first unarmed, the second armed with spiral bands of fine spines

and increasing slightly in width distally. Thread unarmed.

(ii) Large stenoteles, found in capitula of tentacles. Capsule oval, 12,6— -

14,4 x 10,8-12,0 uw. Details of armature not distinguishable.

(111) ?Small stenoteles, abundant in capitula of tentacles and also present

on hydranth body. Capsule oval, 6,0-6,7 x5,0-5,7 u. Not seen

discharged.

Remarks

The nematocysts of this material agree with those described by Philbert

& Weill, except that the two categories of stenoteles are smaller. Warren’s

‘large’ and ‘small’ nematocysts appear to correspond to the large and small

stenoteles respectively.

12 ANNALS OF THE SOUTH AFRICAN MUSEUM

pa} (Gj) a rt) iii i. = IO)

H,J,K ss Op

|) een Ue

A. Sphaerocoryne bedoti. B-C. Coryne ?pusilla, C showing production of regeneration bodies.

D-E. Cladocoryne floccosa. D, hydranth with male gonophores, and E, discharged and undis-

charged macrobasic euryteles. F-L. Zanclea sp. F and G, hydranths from colony commensal

with polyzoan; H and J, large bean-shaped nematocyst and large stenotele from the same

colony; K, hydranth from colony commensal with coral; L, small bean-shaped nematocyst

and large stenotele from the same colony.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 13

Family Corynidae

Coryne ?pusilla Gartner, 1774

rig, 1B, C

Coryne pusilla: Hincks, 1868: 39; pl. 7, fig. 1. Warren, 1908: 289, fig. 4.

Description

Stems unbranched or with a few lateral branches, reaching about 4 mm in

height. Stem closely annulated throughout or with the basal part only roughly

corrugated; perisarc continued over the base of the hydranth as far as the first

tentacles as a very thin transparent membrane, but not expanded in any way.

Hydranths 0,37-0,75 mm in length and about 0,2 mm in maximum diameter.

Tentacles 11-19 in number, 3-5 forming a verticil round the mouth and the

remainder irregularly scattered or with a tendency to form 3-4 alternating

verticils.

Nematocysts: stenoteles of varying size, 8,4 x 4,8-15,0 «9,0 pz.

Remarks

It is impossible to be certain about the identity of this material in the

absence of gonophores. Warren has reported C. pusilla from Natal, and the

annulation of the stem suggests this species. The dimensions and number of

tentacles of the hydranth agree with Warren’s ‘dwarf form’, which, however,

was sterile and thus its identity subject to doubt. The continuation of the perisarc

over the base of the hydranth resembles the condition illustrated by Prévot for

C. fucicola (1959: fig. 7) but it is not swollen or expanded as is the case in

C. muscoides.

Some of the specimens show vigorous asexual reproduction, the tips of

lateral branches becoming separated off as regeneration bodies (Fig. | C).

These bodies (frustules), when completely separate, are very similar to planula

larvae, containing an outer layer of columnar ectoderm and an inner solid

layer of endoderm. They measure 0,41—-0,53 mm in length and 0,13-0,18 mm in

diameter.

Sphaerocoryne bedoti Pictet, 1893

Fig. 1A

Sphaerocoryne bedoti Pictet, 1893: 10; pl. 1, figs 5, 6. Mammen, 1963: 48, figs 16-18.

Clavatella multitentaculata Warren, 1908: 278; pl. 45, figs 7-9.

Sphaerocoryne multitentaculata: Stechow, 1921: 248. Prévot, 1959: 108. Gravier, 1970: 149.

Description

Pedicel (possibly incomplete) and hydranth together reaching a total height

of 3,5 mm. Pedicel increasing in diameter distally, 0,11 mm wide at base, 0,16 mm

14 ANNALS OF THE SOUTH AFRICAN MUSEUM

at distal end; perisarc smooth. Hydranth body 0,7 mm in length (preserved),

with 15 capitate tentacles in two alternating verticils.

Remarks

This is probably a young hydranth.

Family Zancleidae

Zanclea sp.

Fig. 1 F—-L

Description

Four commensal zancleid colonies have been collected in the Inhaca

area, all of which superficially appear to fall within the range of Z. costata, as

discussed by Millard & Bouillon (1973). However, the first of these shows certain

differences in host and cnidome from the other three, and will be described

separately.

Colony: (Fiscal Ker)

Colony commensal with a coral. Hydrorhiza coated with firm perisarc and

ramifying on the surface of the coral skeleton below the soft body. Hydrocaulus

penetrating the coral body and emerging at the surface through a pore, enclosed

in perisarc to just above level of emergence where it becomes membranous and

creased.

Hydranths all infertile, reaching a maximum height of 1,1 mm, with 18-28

stalked capitate tentacles, of which six form an oral whorl and the rest are

irregularly scattered.

No large bean-shaped nematocysts comparable with those of the Seychelles

colonies could be found, but a very few small bean-shaped capsules measuring

13,8 x 6,0 « were seen in the lower part of the body.

Large and small stenoteles, measuring 10,8 x 9,0 and 6,0 x 4,8 » respec-

tively, occur abundantly in the tentacles and the body.

Colonies 2, 3 and 4 (Fig. 1 FJ)

Colonies commensal with polyzoa. Hydrorhiza ramifying within the poly-

zoan and coated with thin perisarec which terminates at the level of emergence

of the hydrocaulus or slightly above.

Hydranths reaching a maximum height of 0,8 mm, with 11-24 tentacles,

of which 3-5 have larger capitulae than the rest and form an oral whorl, and the

remainder are irregularly distributed over the body. Tentacles mostly with

very short stalks and some reduced to mere knobs.

Medusa-buds borne on the lower part of the hydranth either amongst the

lower tentacles or below them, 1-3 per hydranth. Largest medusa 0,16 mm in

diameter and showing rudiments of marginal bulbs.

(hn

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA |

Nematocysts of three types:

(i) Large, bean-shaped capsules, probably macrobasic euryteles, varying

slightly in shape from 19,2 x10,8 » to 22,8 7,8 ». Butt forming at

least one loop in the longitudinal axis. Present in body and hydro-

caulus.

(11) Large stenoteles with almost spherical capsules, reaching 11,4 « 10,8 1,

present in capitulae of tentacles and elsewhere in the body.

(iii) Small stenoteles with alrmost spherical capsules, 6,0 «4,8 u, abundant

in tentacles and also present elsewhere in body.

These colonies show shortening of the tentacles which may be concomitant

with the early reproductive stage. The appearance of the undischarged capsules

of the large bean-shaped nematocysts is slightly different from those of the

Seychelles material described by Millard & Bouillon (1973).

Family Clavidae

Rhizogeton nudum Broch, 1909

Fig. 2A

Rhizogeton nudum Broch, 1909: 137, fig. 1. Ritchie, 1910: 827, fig. 80. Mammen, 1963: 34,

fig. 3. Rees & Thursfield, 1965: 48.

Description

Hydrorhiza reticular, coated with a transparent layer of perisarc.

Hydranths sessile, cylindrical, 0,84-2,00 mm in height, with a very low collar

of perisarc around the base, with 16-26 filiform tentacles scattered over the

distal two-thirds of the body and increasing in length distally so that the most

distal are about three times the length of the most proximal.

Gonophores male, borne directly on hydrorhiza, oval, completely enveloped

in thin perisarc, reaching a maximum size of 0,29 x 0,23 mm.

Remarks

This species was originally described from the Arctic and has since been

reported from the same area several times by Kramp (1932, 1943). Infertile

material has been reported from the Pacific Ocean (Ritchie 1910) and from

India (Mammen 1963). The presence of male gonophores similar to those of —

Broch in the present material seems to confirm the presence of the species in

the Indian Ocean in spite of the peculiar discontinuity of the distribution.

Turritopsis nutricula McCrady, 1856

Turritopsis nutricula: Russell, 1953: 115, figs 54-56; pl. 5, figs 1-5; pl. 29, figs 1-3. Vervoort,

1968: 5. Millard & Bouillon, 1973: fig. 4C.

Description

Larger stems reaching 5 mm in height, branching irregularly, but unfascicled

and presumably still young. Perisarc of stem in two layers, the inner firm and

16 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 2.

A. Rhizogeton nudum, hydranth and male gonophore. B. Podocoryne ?carnea, gastrozooid,

gonozooid and spine. The right gonophore is shown in half section. C. Incertae sedis species 1.

D. Lineolaria sp., surface view of colony with hydrothecae and nematothecae.

corrugated, the outer thin and sometimes with adhering silt. Branches adnate

and parallel to stem for some distance, then diverging at an acute angle. Several |

young and unbranched stems present bearing a single terminal hydranth, only |

slightly more advanced than those illustrated by Russell (1953: pl. 5, fig. 5); |

the perisarc in these also in two layers. ;

Hydranth with 12-38 scattered filiform tentacles, the proximal ones shorter |

than the distal.

Young medusa-buds borne below hydranths on the taller stems, about |

0,10 mm in diameter.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA Vi

Family Eudendriidae

Eudendrium capillare Alder, 1856

Fig. 3 E-H

Eudendrium capillare Alder, 1856: 355; pl. 12, figs 9-12. Picard, 1955: 183. Mammen, 1963:

57, figs 25, 26.

Eudendrium parvum Warren, 1908: 272, fig. 1; pl. 45, figs 1-4.

Eudendrium ?parvum: Millard, 1959: 305, fig. 1 G—-H.

Non Eudendrium ?capillare: Millard, 1966: 454.

Description

Stem unfascicled, unbranched or sparsely and irregularly branched, reaching

1,4 cm in height, annulated or corrugated on origins of branches, hydranth

pedicels often annulated or corrugated throughout. Hydranth with 17-23

tentacles (mean number 19 in eight counts).

Male gonophores present in two colonies, one- to three-chambered, some-

times with a terminal tubercle, borne on completely atrophied hydranths —

even in the youngest there is no sign of tentacles.

Female gonophores present in two colonies; young ones borne by partly

atrophied hydranths in which tentacles are present but reduced in size, spadix

unbranched; one old blastostyle present in which the tentacles are completely

absent and the spadices of the gonophores shed. Eggs orange in colour, reaching

0,22 mm in diameter.

Only small nematocysts present, abundant on tentacles and also present in

hydranth body: microbasic euryteles, 6,6 x 2,4-8,0 x 3,0 uw, undischarged only.

Remarks

This is the first of three small species of Eudendrium which occur at Inhaca.

They are all very similar in their growth-form and all have female gonophores

with an unbranched spadix. As Picard (1951, 1955) has shown, the only sure

way of distinguishing such species is by nematocyst structure. £. capillare

differs from the other two in the complete absence of large nematocysts. It also

differs in the absence of tentacles at all stages of development of the male

blastostyles.

Eudendrium motzkossowskae Picard, i951

Fig. 3:5, K

Eudendrium simplex: Motzkossowska, 1905: 56; pl. 3, figs 18, 19.

Eudendrium motzkossowskae Picard, 1951: 339.

Description

Stems unfascicled, unbranched or sparsely branched, reaching 1,2 cm

in height, smooth for the most part but annulated on origins of branches and

at other irregular intervals, hydranth pedicels sometimes annulated or corru-

18 ANNALS OF THE SOUTH AFRICAN MUSEUM

A-C,E-J

‘ J |mm

Fig. 3.

A-D. Eudendrium ramosum. A, young female blastostyles; B and C, male blastostyles; D, large

microbasic eurytele discharged and undischarged and small microbasic eurytele discharged.

E-H. Eudendrium capillare. E, very young female blastostyle; F, slightly older female blasto-

styles; G, old female blastostyle with spadix shed; H, old and young male blastostyles. J-K.

Eudendrium motzkossowskae. J, male blastostyle; K, large macrobasic eurytele discharged

and undischarged.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 19

gated throughout. Hydranth with 16-27 tentacles (mean number 20 in 11 counts).

Male gonophores present in two colonies, one- and two-chambered, borne

on non-atrophied hydranths. Female gonophores absent.

Nematocysts of two types:

(i) Large macrobasic euryteles present on hypostome and lower part of

hydranth body; sometimes rather scarce and present on only some

hydranths of a colony, at other times abundant. Size rather variable,

15,6 x 6,0-31,6 14,0 yw. Undischarged capsule with butt in 3-4

obliquely longitudinal coils. Discharged capsule with butt 5—6 times

length of capsule (150-165 ju), swollen distally to about double the

width, bearing spirally arranged barbs which appear to be absent

on most of the terminal dilation. Thread less than half width of

proximal part of butt.

(ii) Small microbasic euryteles present on tentacles and elsewhere,

6,6 x 2,4-7,2 3,0 », undischarged only.

Remarks

From the experience of the second author with Mediterranean hydroids,

the size and structure of the large nematocysts in this material indicate an

identity with E. motzkossowskae, so far known only from the Mediterranean.

The discovery of the characteristic female gonophores with hermaphrodite

contents is awaited to completely confirm the diagnosis.

Eudendrium ramosum (Linnaeus, 1758)

Fig. 3 A-D

Eudendrium ramosum: Allman, 1872: 332; pl. 13. Leloup, 1952: 127, fig. 64. Picard, 1955: 183.

Millard, 1966: 456. Millard & Bouillon, 1973: 32, fig. 4F.

Description

Stem unfascicled, unbranched or sparingly branched with an alternate

tendency; almost entirely smooth both on main stem and pediceis except for

groups of a few annulations on origins of branches and at other rare intervals,

reaching 2,0 cm in height. Hydranths with 18-29 tentacles (mean number 23°

in 36 counts).

Male gonophores present in five colonies; one- or two-chambered, borne

on non-atrophied hydranths. Female gonophores present in five colonies,

borne on non-atrophied hydranths, with unbranched spadix.

Nematocysts of two types:

(i) Large microbasic euryteles, present on hypostome and lower part of

hydranth body, on spadix of female gonophore and on distal end of

male gonophore, 22,5 x 10,5-28,8 x 13,8 mw. Undischarged capsule

with butt about three-quarters length of capsule (0,67—0,84), increasing

distally to about twice the diameter. Discharged capsule with butt

20 ANNALS OF THE SOUTH AFRICAN MUSEUM

about three-quarters length of capsule and extruded at right angles

to its main axis, dilated distally to about twice the diameter. Armature

not clearly observed. Occasionally absent in some hydranths of a

colony.

(ii) Small microbasic euryteles, present on tentacles and elsewhere,

6,6 x 2,4-8,4 x4,2 pw. Discharged capsule with butt about three-

quarters length of capsule, dilated in distal half, armature not clearly

observed.

Remarks

This is by far the most common of the three species of Eudendrium. It is

surprising that all the colonies are so small and that none has a fascicled stem.

However, larger colonies reaching 17,5 cm have been reported from the Cape,

South Africa (Millard 1966).

Family Hydractiniidae

Hydractinia diogenes Millard, 1959

Hydractinia diogenes Millard, 1959: 305, fig. 2.

Description

Colonies very similar to holotype. Spines rather scarce, reaching 0,5 mm

in length. Gastrozooids reaching 2,9 mm in height and with 10-16 tentacles.

Gonozooids reaching 1,2 mm in height and with up to nine tentacles. Gono-

phores, all female, of various sizes, the largest 0,45 mm in depth and 0,47 mm

in diameter.

Spiral zooids present on edge of one colony, reaching a length of 2,6 mm

and bearing a terminal battery of nematocysts.

Podocoryne ?carnea M. Sars, 1846

Fig. 2B

Hydractinia carnea: Millard, 1957: 181.

Podocoryne carnea: Edwards, 1972: 122, 136, figs 7-9.

Description

Hydrorhiza forming an open meshwork in younger parts and an incrustra-

tion of coalesced tubes in the older parts. Spines hollow, transparent or horn-

coloured, straight or curved, about 0,3 mm in height. No spiral zooids or tenta-

culozooids.

Gastrozooids reaching 1,14 mm in height (preserved), with 7-10 tentacles,

usually long and short alternating. No perisarcal cup round base.

Gonozooids more slender than gastrozooids and often much smaller,

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 21

reaching 1,14 mm in height (preserved), with 4-7 tentacles, bearing 1-3 gono-

phores on lower half.

Gonophores packed with about 60 eggs in seven or eight tiers, the largest

gonophore 0,57 x 0,49 mm, with four radial canals and a circular canal, and in

some, rudiments of four marginal bulbs or tentacles.

Remarks

These gonophores are very similar to those with precociously developed

sexual products in material ascribed to P. carnea by Millard & Bouillon (1973),

and by Millard (1957). They differ in the larger number of smaller eggs.

Several other infertile hydractiniids on various substrata, including barna-

cles, worm-tubes, empty gastropod shells and a Pecten shell, may possibly be

the same species. Some possess spines and others are without. Some possess

spiral zooids scattered amongst the hydranths.

INCERTAE SEDIS

Species |

Fig: 2€

Description

Hydrocaulus unbranched, perisarc firm and roughly corrugated in lower

part but increasing in diameter and becoming very thin distally to continue

over the body of the hydranth as a membranous layer, terminating below the

tentacles.

Hydranth minute, cylindrical, fairly clearly demarcated from hydrocaulus,

with conical hypostome and a single whorl of 7-9 filiform tentacles held alter-

nately elevated and depressed, no web between the tentacles. Hydranth 0,17-

0,48 mm in length. Total height (hydranth plus hydrocaulus): 0,62—1,16 mm.

Remarks

These specimens show some resemblance to hydranths of the family

Pandeidae, and especially to the genus Amphinema. They resemble A. rugosum

in the corrugated perisarc and A. dinema in the membranous terminal part of ~

the perisarc. It is not possible to be more definite in the absence of gonophores.

Species 2

Description

A stolonial colony with reticular hydrorhiza creeping among sand grains

and giving rise to solitary hydranths. Hydrorhiza and hydrocaulus enclosed in

perisarc. Hydranth emerging from side of hydrocaulus, with 10-13 filiform

tentacles, reaching a maximum height of 0,64 mm when extended. Hydrocaulus

reaching a height of 2,55 mm.

22 ANNALS OF THE SOUTH AFRICAN MUSEUM

Remarks

This is the same species described by Millard & Bouillon (1973: fig. 4H).

The hydrocaulus is taller in this material, but shows no indication of branching.

The systematic position remains uncertain.

Species 3

Lineolaria sp.

Fig. 2D

Lineolaria sp. Gravier, 1970: 144, figs 11, 13A. Gravier, 1972: 8.

Description

Colony growing on Cymodocea. Hydrorhiza generally running parallel

to the long axis of the frond, in single, double or triple strands connected to one

another by cross anastomoses; each longitudinal strand giving rise to a row of

hydrothecae on each side which may be alternate, opposite or irregular.

Hydrotheca tubular, adherent to weed for most of its length, then bent

upwards. Perisarc thickened near base, becoming thin distally. Margin facing

obliquely upwards, circular, untoothed, with an operculum of many delicate

converging segments. Hydranth with about 14 tentacles and a conical

hypostome.

Nematothecae borne on hydrorhiza on the transverse connections, one or

two flanking each hydrotheca and leaning over it, erect, not adherent, tubular,

containing a group of large nematocysts.

Gonothecae absent.

Measurements (mm)

Hydrotheca, length 0 ae Ve ou 2) 042-0557

diameter at margin Be a = - 0,16-0.21

Nematotheca, length on ae ne .. QO, 15=0;22

maximum diameter a, ie ne ie .. 0,04—0,05

Remarks

This appears to be the same species as that briefly described and illustrated

by Gravier from Madagascar. We have refrained from naming the species or

placing it in a family pending a forthcoming publication by Gravier.

Family Haleciidae

Halecium tenellum Hincks, 1861

Halecium tenellum: Vervoort, 1959: 229, fig. 8. Millard, 1966: 471, fig. 11 C—F.

Description

This small species occurs abundantly on any available substratum. Some of

the stems are beautifully regular and geniculate, with a hydrotheca arising from

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 23

each internode. Others are less regular, with numbers of intervening athecate

internodes, particularly in the basal region. Branches arise singly or in pairs

from the distal ends of the internodes. Rejuvenation is common and many

tiers of secondary hydrothecae occur.

Male and female gonothecae occur on separate colonies and arise either

from the hydrorhiza or from the upright stem. They are similar to those pre-

viously described (Millard 1966).

Hydrodendron gardineri (Jarvis, 1922)

Fig. 4

Halecium gardineri Jarvis, 1922: 334; pl. 24, fig. 1.

Description

Hydrorhiza reticular and strengthened by internal thickenings of perisarc,

giving rise to solitary hydrophores and erect stems. Stem reaching a maximum

height of 4 mm, unfascicled, unbranched, divided into internodes by oblique

nodes sloping in alternate directions. Each internode bearing a single hydrophore

on a short apophysis arising near the centre or at about two-thirds of the length.

Hydrophores alternate, the two rows in one plane.

Hydrophore (Fig. 4B) consisting of a pedicel separated from the apophysis

by a twisted node, bulging slightly at base, then expanding smoothly to merge

into the cup-shaped hydrotheca; completely symmetrical or with an oblique

margin. Hydrotheca deep, with oblique diaphragm sloping downwards on

adcauline side; margin not, or slightly, everted; marginal diameter

approximately equal to depth on adcauline side. Many hydrophores regene-

rated, resulting in supplementary internodes, which are often roughly corru-

gated, between the apophysis and the pedicel (Fig. 4A). Solitary hydrophores

usually with a number of supplementary internodes at base (Fig. 4C and D).

No refringent nodules in hydrotheca.

Nematothecae borne irregularly on hydrorhiza, on stem and on regeneration

nodes, but none seen on unregenerated pedicels, goblet-shaped, with everted

margin. Nematophore with a solid core of endoderm cells and a terminal

capitulum.

Hydranth with 18 tentacles in the only expanded example, and no inter-

tentacular web, large but apparently just able to be contained in the hydrotheca.

Nematocysts of at least two types (Fig. 4G):

(i) Small microbasic mastigophores, abundant on tentacles. Capsule

5,4x1,8 pw (discharged). Butt 4,8 » in length, with about six spiral

bands of short spines.

(ii) Large heteronemes, seen on nematophores only, 18,0x6,6 yu

(undischarged).

Gonophores absent.

24 ANNALS OF THE SOUTH AFRICAN MUSEUM

toMy

Fig. 4. Hydrodendron gardineri.

A-B. Parts of erect stem showing hydrophores and nematophores. C—D. Solitary hydrophores,

D with a nematotheca. E-F. Parts of an erect stem from Jarvis’s schizotype. G. Macrobasic

mastigophore and large heteroneme.

Measurements (mm)

Jarvis’s type

Stem, internode length 0,18—0,34 0,30-0,50

diameter across node a. .. 0,06—0,09 0,08—-0,10

Hydrophore, length adcauline, unregenerated 0,20-0,23 0,24—0,30

Hydrotheca, depth adcauline 0,11-0,16 0,11-0,14

diameter at margin 0,13-0,17 0,14-0,17

Nematotheca, depth 0,08—0,10 0,08—0,09

diameter at margin 0,05—0,07 0,05

Remarks

Through the courtesy of the British Museum (Natural History) we were

able to examine a prepared slide of Jarvis’s type material of Halecium gardineri

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 25

(Fig. 4 E and F: schizotype, no. 23.2.15.9) from Chagos and there is no doubt

about the identity of the Inhaca material. Measurements from the type are

included above as none were given by Jarvis and her quoted magnification is

not accurate.

Hydrodendron sympodiformis n. sp.

Fig:

Holotype: a very rich fertile colony on weed from Inhaca, Ponta Torres.

Description

Hydrorhiza creeping, branching and reticular, with internal thickenings of

perisarc in some areas only.

Hydrophores arising from hydrorhiza, either unbranched and solitary or

branching in a very obviously sympodial manner to reach a maximum height

of 7 mm. Solitary hydrophores pedicellate, with pedicel of variable length

widening distally to merge into hydrotheca. Shorter pedicels completely smooth,

longer ones (possibly regenerated) with one or two nodes or corrugations near

base. In branching stems each hydrophore giving off a lateral apophysis near

distal end which bears the next hydrophore, this process being repeated up to

nine times, the base of each hydrophore being curved upwards so that it stands

almost parallel to the preceding one; the series of apophyses not in one plane

but arising from any surface. Two apophyses sometimes arising at the same

level, so producing a dichotomous appearance.

Hydrotheca widening to margin, which may or may not be everted, with

diameter usually a little over twice the depth, with a straight diaphragm and a

circle of refringent nodules between margin and diaphragm. Hydranth with

about 22 tentacles.

Nematothecae borne in random fashion on hydrorhiza and hydrophore

pedicels, vase-shaped, with an everted margin and a perisarcal thickening just

below it. Nematophore with a solid core of endoderm cells and a terminal

capitulum containing a battery of large nematocysts about 22,5 « 10 wu.

Gonothecae borne in abundance on hydrorhiza, all female, barrel-shaped

but with widest part below centre, with 6—7 very distinct transverse annulations

and a terminal aperture on a narrowed neck, containing one planula larva.

Measurements (mm)

Hydrotheca, depth, margin to ee re .. 0,06—0,10

diameter at margin i Ae e. ne. (0507-022

Nematotheca, depth as a ae yh .. 0,14-0,20

diameter at margin a a Me - .. 0,09-0,11

Gonotheca, depth .. st es ae as .. 0,63-0,80

maximum diameter ae i - ng .. 0,39-0,44

diameter at margin i a: a be .. 0,20-0,26

26 ANNALS OF THE SOUTH AFRICAN MUSEUM

A BaD

a a et || inition] a ee a eee orm imine

Fig. 5. Hydrodendron sympodiformis n. sp.

A. Complete stem and gonotheca. B. Solitary hydrophore and nematothecae. C. Part of

stem. D. Gonotheca.

Remarks

The growth-form of this material shows similarities to that of H. negligens

(Fraser 1938) and H. alternata (Fraser 1938) from the Pacific, but certain

differences from the structure of the trophosome of both and from the shape

of the gonotheca in the latter seem to merit the establishment of a new species.

Family Campanulariidae

Campanularia delicata (Trebilcock, 1928)

Orthopyxis delicata Trebiicock, 1928: 3; pl. 2, fig. 1. Ralph, 1957: 837, 840, fig. 7 a-d.

Campanularia delicata: Millard & Bouillon, 1973, fig. 6 G—-M.

Remarks

The hydrothecae and gonothecae (male) of this material are similar to

those described by Millard & Bouillon (1973), from the Seychelles and show

similar variations. The hydrothecal pedicels in particular show much variation

in length and amount of annulation.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA pay |

Clytia sp.

Fig. 6A

Description

Colony stolonial. Pedicel closely annulated at base and distal end. Hydro-

theca deep and slender, with height 24-33 times diameter, margin with 5-8

triangular, slightly asymmetrical teeth. Diaphragm well developed.

Measurements (mm)

Pedicel, height - BY cee ins ae .. 0,44-1,12

Hydrotheca, depth .. red a ce ae .. 0,48-0,59

diameter at margin xe w x) ie ao, (O14=0.24

Remarks

The hydrothecae have the same proportions as C. warreni Stechow, 1919,

from South Africa, C. e/ongata Marktanner-Turneretscher, 1890, from New

Zealand, and C. ulva Stechow, 1919, from Marseille. They differ from

C. elongata in the deeper marginal teeth. The actual size is intermediate between

that of C. ulvae and C. warreni. The marginal teeth are fewer in number than

any of the three species mentioned.

In the absence of gonothecae and because of the small size of the sample

it is not possible to diagnose the material with any certainty. It might possibly

be an extreme variation of C. hemispnaerica.

Family Lafoeidae

Scandia mutabilis (Ritchie, 1907)

Description

Hydrothecal pedicels very variable in length, 0,3-4,2 mm, but always

annulated. Hydrothecae also variable in size, 1,6-3,9 mm in height, generally

quite smooth, but a few with very faint transverse corrugations.

Family Sertulariidae

Abietinaria laevimarginata (Ritchie, 1907)

Fig. 7 A-C, E-H

Sertularia laevimarginata Ritchie, 1907: 507; pl. 26, figs 5, 6.

Sertularia linealis Warren, 1908: 308, fig. 9. Millard, 1958: 195, fig. 8 D, G.

Abietinaria laevimarginata: Stechow, 1921: 258. Gravier, 1972: 8, fig. 2C.

Non Sertularia linealis var. longa Millard, 1958: 197, fig. 8E. Millard & Bouillon, 1973, fig.

OE. FE.

Non Sertularia linealis: Millard, 1968: 272.

Description

Hydrorhiza creeping on weed and typically forming longitudinal lines

with cross-connections; with no regular internal pegs of perisarc, but usually

28 ANNALS OF THE SOUTH AFRICAN MUSEUM

\{/

C-E

0,8 mm

Fig. 6.

A. Clytia sp. B. Diphasia digitalis, unbranched stem. C. Diphasia tetraglochina, unbranched

stem. D. Dynamena crisioides, var. alternata, hydrocladium. E. Sertularia marginata, part of

stem and hydrocladium.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 29

A, E-F. Sertularia laevimarginata Ritchie, holotype (E and F from different stems). B, G. Ser-

tularia linealis Warren, holotype. C, H. Abietinaria laevimarginata, from Inhaca. D, J. Sertularia

longa, from Inhaca. (In E-J the distal part of the hydrotheca is drawn without the near wall

to show details of internal teeth and operculum.)

1mm

Ce ggg ge dO? mim

30 ANNALS OF THE SOUTH AFRICAN MUSEUM

with a group of four ingrowing perisarcal lobes around the origin of each stem.

Stem reaching a maximum height of 3,5 mm and bearing a pair of opposite

hydrothecae on each internode. Two hinge-joints at the base of each stem,

remaining nodes slightly oblique or indistinct.

Hydrotheca bent outwards, narrowing to margin and usually constricted

just below it, with thickened perisarc, especially near the adcauline edge. The

free adcauline parts of a pair of hydrothecae forming a straight line at right

angles to the axis of the stem, except occasionally in the distal pair which are

more erect and may subtend an obtuse angle between them. Margin with two

low, rounded lateral lobes, which vary considerably in development —they may

be quite distinct or almost obsolete. Operculum of one adcauline valve, shed

easily. Internal teeth present, including one large adcauline tooth (occasionally

double) and two low latero-abcauline teeth.

Gonotheca borne on front of stem below first pair of hydrothecae, smooth,

compressed, spherical to ovoid in broad view, with a wide distal aperture on a

low collar.

Remarks

Gravier (1972) has synonymized Warren’s Sertu/aria linealis with Abietinaria

laevimarginata. In order to confirm this course of action, and to clarify Warren’s

statement that S. /inealis has a ‘two-flapped operculum’, the first author has

examined type material of Warren’s S. /inealis borrowed from the Natal Museum,

Pietermaritzburg (No. 747) and from the British Museum of Natural History

(Nos. 22.3.6.26, 27 and 33, Fig. 7 B, G). In both these collections whole mounts

clearly show a single adcauline opercular valve. But also included in the British

Museum material is a slide of sections (No. 22.3.6.32) of a completely different

species, with a hydrotheca of a different shape and a two-valved operculum

(labelled S. /inealis but probably S. distans). This confusion might thus account

for Warren’s erroneous statement.

The schizoholotype of S. /aevimarginata Ritchie (B.M. No. 1964.8.7.148)

consists of two infertile stems (Fig. 7 A, E, F). The perisarc throughout is

somewhat thinner than in S. /inealis, and internal thecal teeth are generally

absent, though in one hydrotheca three internal teeth are visible, one adcauline

and two low latero-abcauline. Around the origin of each stem is one ingrowing

lobe of perisarc, not four as in S. Jinealis; however, even in the latter species

these lobes are not always conspicuous. The details of shape and dimensions

of the hydrothecae are sufficiently like those of S. Jinealis to confirm the syno-

nymy in spite of the minor differences.

With the abundant new material it has now become clear that the first

author has previously confused two species under the name of S. Jinealis, due

partly to the fact that both have a similar growth-form on weed and may grow

together, and to the fact that the hydrothecae of A. laevimarginata easily lose

the opercula. The second form, described by Millard (1958) as Sertularia linealis

var. /onga is now elevated to specific rank (see p. 33 and Fig. 7 D, J). The two

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 3]

species can be distinguished not only by the form of the operculum, but by

differences in the hydrorhiza, presence or absence of internal teeth, and the

shape of the gonotheca.

Diphasia digitalis (Busk, 1852)

Fig 6B

Desmoscyphus longitheca Allman, 1877: 26; pl. 14, figs 3-6.

Nigellastrum digitale: Mammen, 1965: 57, fig. 89.

Diphasia digitalis: Vervoort, 1968: 37, fig. 17. Millard & Bouillon, 1973, fig. 9A.

Description

Stems either unbranched, or with alternate branches arising after every two

or three pairs of hydrothecae. Branches forming almost a right angle with stem.

Hydrothecae in opposite pairs; members of a pair well-separated near the base

of the stem, but shifted on to the front and contiguous with one another in the

distal parts and on the branches; consecutive pairs close, separated by a distance

not more than 3 of their height, but more often in contact or overlapping.

Hydrotheca tubular and curved outwards, usually with three or five longi-

tudinal ridges, in the latter case two anterior, two posterior and one lateral.

Gonothecae absent.

Remarks

The branching of this material is very similar to that figured by Allman

(1877) (as Desmoscyphus longitheca). In some stems the branches are similarly

short, in others they are longer.

The hydrothecae are similar to those described from the Seychelles (Millard

& Bouillon 1973), but the longitudinal ridges are more distinct.

Diphasia tetraglochina Billard, 1907

Fig. 6C

Diphasia tetraglochina Billard, 1907: 358, fig. 7. Millard, 1964: 28, fig. 8.

Description

Two of these colonies are exactly similar to the material from the Agulhas

Bank, South Africa (Millard 1964), i.e. the hydrothecae are shorter and wider

but with a greater proportion adnate (2-3) than those of the type material

(Billard 1907).

The third colony, from Cabo da Inhaca (Fig. 6C), is more like Billard’s

material with taller and more slender hydrothecae with a smaller proportion

adnate (4-3).

32 ANNALS OF THE SOUTH AFRICAN MUSEUM

Measurements of material from Cabo da Inhaca, in mm.

Internode length .. bil , bs a .. 0,52-0,63

Hydrotheca, length abcauline wm a 1. OA0=0s1

length adcauline, adnate part .. #3 5 .., 0,31—-0;39

length adcauline, free part a: is oe .. 0,18-0,26

adnate part/adcaulinelength .. ie 5. .. 0,57-0,64

diameter at margin sts Bec WA te .»/ OF15-0:20

Dynamena crisioides Lamouroux, 1824

Fig. 6 D

Dynamena crisioides: Billard, 1925: 181, figs 36, 37 C—E; pl. 7, fig. 21. Millard, 1958: 183.

Mammen, 1965: 51, figs 84, 85.

Dynamena crisioides, var. peculiaris Billard, 1925: 185, fig. 38.

Dynamena crisioides, var. gigantea Billard, 1925: 186, fig. 37F; pl. 8, fig. 24. Millard, 1958:

183, fig. 6C.

Dynamena crisioides, var. alternata Billard, 1925: 187, figs 37G, 39; pl. 7, fig. 22.

Description and remarks

This species is abundant in the collection and shows great variation. The

normal variety and Billard’s var. gigantea both occur and sometimes in the

same colony. Possibly the latter is an older stage of the former.

One sample of Billard’s var. alternata is also present (Fig. 6D), which

shows the characters described by Billard, namely two alternate hydrotheca

on most of the hydrocladial internodes, hydrothecae less adnate than in the

typical form, and hydrocladial apophyses arising below the third hydrotheca

of a stem internode instead of below the first as in the typical form. The hydro-

thecae are smaller than those of the typical form, measuring 0,25—0,33 mm in

abcauline length as against 0,34-0,51 mm.

Internal teeth may occur in hydrothecae of any of the three forms, as in

Billard’s var. peculiaris.

Without doubt Mammen is right in his contention that it is not possible

to retain separate varieties in this very variable species.

Sertularella

In addition to the single species listed on p. 8, several small specimens of

Sertularella are present in the collection, but all are infertile and all appear

to be juvenile colonies. Since it is not possible to be certain of the form of the

adult colony in any of them, no attempt has been made at identification. There

appear to be about three species involved.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 33

Sertularia longa (Millard, 1958)

Fig. 7 D and J

Sertularia linealis, var. longa Millard, 1958: 197, fig. 8E.

Sertularia linealis: Millard, 1968: 272.

Sertularia linealis longa: Millard & Bouillon, 1973, fig. 9 E, F.

Description

Hydrorhiza creeping on weed and usually forming longitudinal lines with

cross-connections ; with numerous strengthening internal pegs of perisarc growing

in from the sides.

Stem reaching about 3 mm in height and bearing a pair of opposite hydro-

thecae on each normal internode. An oblique hinge-joint (or rarely two) at

the base of each stem and hinge-joints occurring sporadically in the rest of the

stem where each one forms the distal termination of a short intermediate

athecate internode. Normal nodes slightly oblique.

Hydrotheca bent outwards, and abcauline wall with a definite kink in the

lower or middle part, narrowing to margin, with comparatively thin perisarc

but thickened around margin. The free adcauline parts of a pair of hydrothecae

not forming a straight line, but subtending an obtuse angle between them which

becomes smaller near the distal end of the stem where the hydrothecae are

more erect. Margin with two well-developed and roundly triangular lateral

teeth and sometimes a very low median adcauline one. Operculum of two valves

hinged at the adcauline and abcauline edges, the adcauline smaller than the

abcauline and divided into two by a median line. No internal teeth.

Remarks

This species has hitherto been confused with Abietinaria laevimarginata.

See remarks on this species.

Sertularia marginata (Kirchenpauer, 1864)

Fig. 6E

Sertularia marginata: Millard, 1957: 224, fig. 13. Ralph, 1961: 785, fig. 12 a-g. Van Gemerden-.

Hoogeveen, 1965: 39, figs 13-17. Mammen, 1965: 45, fig. 77.

Sertularia inflata: Vervoort, 1959: 281, figs 39-41. Van Germerden-Hoogeveen, 1965: 45, figs

18-22.

Description

Pinnate stems, on the whole with very regular segmentation, with few

internodes which do not bear the normal hydrocladium and three hydrothecae.

Basal athecate part of stem variable in length, 0,7-3,0 mm, terminated by a

hinge-joint and containing a variable number of transverse nodes. Between

hinge-joint and first hydrocladium 0-2 hydrothecae. Stem geniculate in distal

region only. The two rows of hydrocladia in one plane.

34 ANNALS OF THE SOUTH AFRICAN MUSEUM

Hydrocladium unbranched, separated from basal apophysis by hinge-

joint and thereafter divided by very indistinct nodes into thecate internodes.

Pairs of hydrothecae more closely set than in the South African material from

False Bay (Millard 1957), the interval between consecutive pairs being usually

less than, and never more than, the height of the hydrothecae. Hydrothecae also

smaller and with less pronounced marginal teeth, 0,15—0,22 mm in abcauline

height and 0,09-0,12 mm in marginal diameter.

Remarks

Among recent authors opinions differ as to whether or not S. marginata

and S. inflata should be retained as separate species. Van Gemerden-Hoogeveen

claims to be able to distinguish between the two in Caribbean material, but the

present material, and also that of Mammen (1965) and Ralph (1961), seems to

combine features of both. The present authors follow the opinion first expressed

by Billard (1925: 178) that Sertularia marginata, S. flosculus, S. amplectens,

S. gracilis, S. inflatus and S. versluysi are all synonymous.

Family Plumulariidae

Plumularia obliqua (Johnston, 1847)

Fig. 8 A-D

Plumularia obliqua: Bale, 1884: 138; pl. 12, figs 1-3. Hincks, 1868: 304, fig. 36; pl. 67, fig. 1.

Pennycuik, 1959: 180.

?Monotheca posidoniae Picard, 1951: 341, fig. 2B.

Description

Details of structure conforming weil with previous descriptions. Thickness

of perisarc variable, young hydrociadia with no internodal septa behind hydro-

theca, old ones with two strong septa.

Hydrotheca with an abcauline wall which is gently curved throughout or

straight in the distal half.

Lateral nematothecae with the upper (abcauline) surface strongly cut

away, as described by Bale.

Measurements (mm)

Stem, internode length Hae ah a of .. 0,26-0,44

Hydrocladium, thecate internode, length - 4.4 10, 172028

Hydrotheca, abcauline height we ae ay .. O7021

diameter at margin by: i a by! .. 0,14-0,16

Remarks

Picard (1951) has described a new species from the Mediterranean, P.

posidoniae, which is said to differ from P. obliqua in the deeper hydrotheca,

more convex abcauline wall and better developed perisarc. He includes Bale’s

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 35

ee sO 4m eee NC AD

Fig. 8.

A-D. Plumularia obliqua. A and B, hydrocladia in side view, A with well-marked internodal

septa, B without; C, hydrocladium in posterior view; D, lateral nematotheca. All from the

same colony. E-F. Aglaophenia cupressina. E, part of a hydrocladium; F, the proximal part

of two corbula ribs from the side.

material from Australia in this species. The present material shows sufficient

variation to cover both and it is felt that there is little justification for a separate

species. Picard gives no measurements and does not describe the structure of the

lateral nematothecae.

This is the first record of the species from the East African coast.

36 ANNALS OF THE SOUTH AFRICAN MUSEUM

Aglaophenia cupressina Lamouroux, 1816

Fig: 8 Bo

Aglaophenia cupressina Lamouroux, 1816: 169. Billard, 1913: 107, fig. 96; pl. 6. Bale, 1915:

319; pl. 47, figs 6-8.

Aglaophenia macgillivrayi: Allman, 1883: 34; pl. 10; pl. 20, figs 4-6.

Description

Several robust colonies of this beautiful species. Main stem branching and

rebranching irregularly up to the fourth order. Final branches (pinnae) arising

at regular intervals in subopposite pairs, bearing alternate hydrocladia. Hydro-

cladia also present on more distal parts of stem and larger branches.

Details of hydrocladia, hydrothecae and nematothecae as in previous

descriptions. The slight longitudinal ridge down the centre of the lateral surface

of the hydrotheca mentioned by Bale visible in macerated specimens only.

Hydrotheca 0,25-0,28 mm in depth and 0,14-0,15 mm in marginal diameter.

Margin smooth or sinuated, often with two or three pairs of low rounded lobes.

Corbulae abundant, replacing hydrocladia, with a pedicel of one thecate

segment and about six pairs of ribs, completely closed, reaching 2,8 mm in

length and 1,1 mm in diameter. The first two or three nematothecae of the rib

seated on a raised lobe. Some corbulae with a free branch to the first rib of one

side.

Remarks

This well-known Indo-Pacific species has been reported from Zanzibar

on the East African coast by Kirchenpauer (1872, as A. spicata) and by Jarvis

(1922). This is the most southerly record for the African coast.

Gymnangium gracilicaule lignosum (Millard, 1968)

Halicornaria gracilicaulis lignosa Millard, 1968: 283.

Remarks

These stems are not so woody or thick as is typical of this subspecies and

most have only two orders of branching, one alone having three orders. They

are probably young colonies.

The hydrotheca, however, is of the typical shape, 1.e. short (about 0,23 mm

in total height), with a pronounced sigmoidal curvature to the abcauline wall

and a short distance (about 0,07 mm) between the abcauline margin and the

point of separation of the median inferior nematotheca from the hydrotheca.

The median inferior nematotheca is short and does not reach the level of the

thecal margin.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 37

Family Proboscidactylidae

Proboscidactyla (Lar) sp.

Fig. 9

Description

Colonies growing on tubes of the polychaet, Laonome sp. Hydrorhiza

reticular, spreading between the sand-grains of the polychaet tube. Gastrozooids

forming a single row on rim of tube, reaching a maximum height (preserved)

of 0,48 mm to tip of hypostome. Two solid filiform tentacles arising at half to

two-thirds of height and facing cavity of polychaet tube. Hypostome with an

asymmetrically placed pad of nematocysts on the side away from the cavity.

Gonozooids of varying size, the largest ones reaching 0,31 mm in height

and seated immediately below the gastrozooids or arising from their bases, only

these large gonozooids bearing medusa-buds. Smaller, sterile gonozooids

scattered for some 2 mm down the length of the polychaet tube. Gonozooids

without tentacles, with a distal pad of nematocysts, bearing up to four medusae

of different ages.

0.3 mm

Fig. 9. Proboscidactyla sp.

A. Gastrozooid. B. Sterile gonozooid. C. Fertile gonozooid with medusa-buds. D. An older

medusa-bud.

Medusa with four perradial marginal bulbs each bearing a hollow tentacle,

four interradial nematocyst sacs (cnidothylacies), four unbranched radial

canals, stomach with the beginnings of four pouches, no ocelli. Largest medusa

0,23 mm in depth and 0,26 mm in diameter.

Nematocysts of three types:

(i) Macrobasic euryteles, 13,2-15,6 x 6,0-6,6 . Butt coiled in four whorls

in the transverse axis. Discharged capsules similar to those illustrated

38 ANNALS OF THE SOUTH AFRICAN MUSEUM

by Russell, 1938, for Lar sabellarum. Present in the nematocyst pads

of the hydranths and in the cnidothylacies of the medusae.

(11) Microbasic euryteles, 5,4-7,8 x 2,1-3,0 ». None seen discharged.

(iii) Desmonemes, 3,6-4,2 x 2,4-3,0 wu.

Remarks

Identification of species of Proboscidactyla essentially depends on details

of medusa structure, and as far as the hydranths go ‘there appears to be no

precise way to separate the various species’ (Calder 1970). Hand (1954) advocates

the use of small differences in nematocysts for this purpose, but this is of little

help in the present material, for although the three catagories characteristic

of the genus are present, all are smaller than those described for any other

species.

The hydranths of five species of Proboscidactyla are known, namely

P. circumsabella Hand, 1954, P. flavicirrata Brandt, 1835, P. occidentalis (Fewkes,

1889), P. ornata (McCrady, 1857) and P. stellata (Forbes, 1846) = Lar sabellarum

(Gosse, 1857). Of these, only P. ornata has been reported from the western

Indian Ocean (Kramp 1965: near Mombasa). The life-history of this species

has been described by Brinckmann & Vannucci (1965) and by Calder (1970)

There is nothing to debar the present material from inclusion therein apart from

nematocyst size, and nematocyst size must be variable for there are differences

between the measurements given for this species by Brinckmann & Vannucci

from the Mediterranean and those given by Calder from Virginia. However,

it would be unwise to make a definite diagnosis until such time as the medusa

has been reared.

SUMMARY

A collection of hydroids from the Inhaca area and from Santa Carolina

on the Mocambique coast, East Africa, is described. The collection includes 86

species, of which one is new to science, namely Hydrodendron sympodiformis, and

19 are new records for Africa south of 20° S. Lat.

ACKNOWLEDGEMENTS

The authors wish to thank the following: Professor J. H. Day, University

of Cape Town, for the identification of a polychaet, the Coelenterate Section

of the British Museum of Natural History for the loan of material from the

collections of Warren, Ritchie and Jarvis, and the Natal Museum, Pieter-

maritzburg, for the loan of Warren’s holotype of Sertu/aria linealis.

REFERENCES

ALDER, J. 1856. A notice of some new genera and species of British hydroid zoophytes.

Ann. Mag. nat. Hist. (2) 18: 353-362.

ALLMAN, G. J. 1872. A monograph of the gymnoblastic or tubularian hydroids. Part 2. Ray

Soc. Publs.

A COLLECTION OF HYDROIDS FROM MOCAMBIQUE, EAST AFRICA 39

ALLMAN, G. J. 1877. Report on the Hydroida collected during the exploration of the Gulf

Stream by L. F. de Pourtalés, assistant United States coast survey. Mem. Mus. comp.

Zool. Harv. 5: 1-66.

ALLMAN, G. J. 1883. Report on the Hydroida dredged by H.M.S. Challenger during the years

1873-76. Part 1. Plumularidae. Rep. Voy. Challenger 1873-76 7 (20): 1-54.

BALE, W. M. 1884. Catalogue of the Australian hydroid zoophytes. Sydney: Australian Museum.

BALE, W. M. 1915. Report on the Hydroida collected in the Great Australian Bight and other

localities. III. Zool. Results Fish. Exp. ‘Endeavour’ 3: 241-336.

BEHNER, A. 1914. Beitrag zur Kenntnis der Hydromedusen. Z. wiss. Zool. 3: 381-427.

BILLARD, A. 1907. Hydroides de Madagascar et du sud-est de l’Afrique. Archs Zool. exp. gén.

(4) 7: 335-396.

BILLARD, A. 1913. Les hydroides de l’expédition du Siboga. I. Plumulariidae. Siboga Exped.

Ja: 1-115.

BILLARD, A. 1925. Les hydroides de l’expédition du Siboga. II. Synthecidae et Sertularidae.

Siboga Exped. 7b: 117-232.

BRINCKMANN, A. & VANNUCCI, M. 1965. On the life-cycle of Proboscidactyla ornata (Hydro-

medusae, Proboscidactylidae). Pubbl. Staz. zool. Napoli 34: 357-365.

BRINCKMANN-Voss, A. 1970. Anthomedusae/Athecatae (Hydrozoa, Cnidaria) of the Medi-

terranean. Part 1. Capitata. Fauna Flora Golfo Napoli 39: 1-96.

Brocu, H. 1909. Die Hydroiden der arktischen Meere. Fauna arct. 5: 129-248.

CALDER, D. R. 1970. North American record of the hydroid Proboscidactyla ornata (Hydrozoa,

Proboscidactylidae). Chesapeake Sci. 11: 130-132.

Epwarps, C. 1972. The hydroids and the medusae Podocoryne areolata, P. borealis and

P. carnea. J. mar. biol. Ass. U. K. 52: 97-144.

Gravier, N. 1970. Etude des Hydraires epiphytes des Phanérogames marines de la région de

Tulear (Sud-oest de Madagascar). Rec/ Trav. Stn mar. Endoume-Marseille (hors série)

10: 111-161.

GrRAvieR, N. 1972. Hydroides 4piphytes de trois phanérogames marines en provenance de

Nossi-bé (N.W. de Madagascar). Téthys Supp!. 3: 3-10.

HAND, C. 1954. Three Pacific species of ‘Lar’ (including a new species), their hosts, medusae,

and relationships. (Coelenterata, Hydrozoa). Pacif. Sci. 8: 51-67.

Hincks, T. 1868. A history of the British hydroid zoophytes. London: Van Voorst.

JARVIS, F. E. 1922. The hydroids from the Chagos, Seychelles and other islands and from the

coasts of British East Africa and Zanzibar. Trans. Linn. Soc. Lond. (Zool.) 18: 331-360.

KIRCHENPAUER, G. H. 1872. Ueber die Hydroidenfamilie Plumularidae, einzelne Gruppen

derselben und ihre Fruchtbehdlter, I. Aglaophenia, Lx. Abh. Naturw. Hamburg 5: 1-52.

KraAmp, P. L. 1932. Hydroids collected in West-Greenland fjords in 1911 and 1912. Meddr

Grénland 91 (3): 1-34.

KRAMpP, P. L. 1943. The zoology of East Greenland. Hydroida. Meddr Grénland 121 (11): 1-52.

KRamp, P. L. 1965. The Hydromedusae of the Pacific and Indian Oceans. Dana Rep. 63: 1-161.

LAMOUROUX, J. V. F. 1816. Histoire des polypiers coralligénes flexibles, vulgairement nommeés

zoophytes. Caen: Poisson.

LeLoup, E. 1952. Faune de Belgique. Coelentérés. Bruxelles.

MAcNaAE, W. & KALK, M. 1958. A natural history of Inhaca Island, Mogambique. Johannesburg:

Witwatersrand University Press. ;

MAmMMEN, T. A. 1963. On a collection of hydroids from South India. I. Suborder Athecata.

J. mar. biol. Ass. India 5: 27-61.

MaAmMEN, T. A. 1965. On a collection of hydroids from South India. II. Suborder Thecata

(excluding Family Plumulariidae). J. mar. biol. Ass. India 7: 1-57.

MILLARD, N. A. H. 1957. The Hydrozoa of False Bay, South Africa. Ann. S. Afr. Mus. 43:

173-243.

MILLARD, N. A. H. 1958. Hydrozoa from the coasts of Natal and Portuguese East Africa.

Part I. Calyptoblastea. Ann. S. Afr. Mus. 44: 165-226.

MILLARD, N. A. H. 1959. Hydrozoa from the coasts of Natal and Portuguese East Africa.

Part II. Gymnoblastea. Ann. S. Afr. Mus. 44: 297-313.

MILLARD, N. A. H. 1964. The Hydrozoa of the south and west coasts of South Africa. Part II.

The Lafoeidae, Syntheciidae and Sertulariidae. Ann. S. Afr. Mus. 48: 1-56.

MILLARD, N. A. H. 1966. The Hydrozoa of the south and west coasts of South Africa. Part III.

The Gymnoblastea and small families of Calyptoblastea. Ann. S. Afr. Mus. 48: 426-487.

40 ANNALS OF THE SOUTH AFRICAN MUSEUM

MILLARD, N. A. H. 1968. South African hydroids from Dr Th. Mortensen’s Java—South

Africa Expedition, 1929-1930. Vidensk. Meddr dansk naturh. Foren. 131: 251-288.

MILLARD, N. A. H. & BOUILLON, J. 1973. Hydroids from the Seychelles (Coelenterata).

Annls Mus. r. Afr. cent. Sér Svo (Sci. zool.) 206: 1-106.

Mortz-KossowskA, S. 1905. Contribution a la connaissance des Hydraires de la Méditerranée

occidentale. |.— Hydraires. Gymnoblastiques. Archs Zool. exp. gén. (4) 3: 39-98.

PENNYCUIK, P. R. 1959. Faunistic records from Queensland. Part V. Marine and brackish

water hydroids. Pap. Dep. Zool. Univ. Qd 1: 141-210.

PHILBERT, M. 1936. Etudes sur Cladocoryne floccosa Rotch. Bull. Inst. océanogr. Monaco

708: 1-16.

PICARD, J. 1951. Note sur les hydraires littoraux de Banyuls-sur-Mer. Vie Milieu 2: 338-349.

Picarb, J. 1955. Hydraires des environs de Castiglione (Algérie). Bull. Stn Aquic. Péche

Castiglione (n.s.) 7: 181-199.

Pictet, C. 1893. Etude sur les hydraires de la Baie d’Amboine. Revue suisse Zool. 1: 1-64.

Prevot, E. 1959. Morphologie et evolution des structures tentaculaires chez les hydraires

gymnoblastes Capitata. Recl Trav. Stn mar. Endoume (Bull.) 17 (fasc. 29): 91-126.

RAcPH, P. M. 1957. New Zealand thecate hydroids. Part I. Campanulariidae and Campanu-

linidae. Trans. R. Soc. N.Z. 84: 811-854.

RALPH, P. M. 1961. New Zealand thecate hydroids. Part III. Family Sertulariidae. Trans. R.

Soc. N.Z. 88: 749-838.

Rees, W. J. & Thursfield, S. 1965. The hydroid collections of James Ritchie. Proc. R. Soc.

Edinb. (B) 69: 34-220.

Ritcuie, J. 1907. On collections of the Cape Verde Islands marine fauna, made by Cyril

Crossland, M.A. (Cantab.), B.Sc. (Lond.), F.Z.S., of St Andrews University, July to

September, 1904. Proc. zool. Soc. Lond. 1907: 488-514.

Ritcuie, J. 1910. Hydroids from Christmas Island, Indian Ocean, collected by C. W. Andrews.

D:Se:, F-R-S: E-Z:S.. ml908: Brocszool) Soc. Lond. 1910" O)826-s3o

RUSSELL, F. S. 1938. On the nematocysts of Hydromedusae. J. mar. biol. Ass. U. K. 23: 145-165.

RUSSELL, F. S. 1953. The medusae of the British Isles. Cambridge: University Press.

STECHOW, E. 1921. Neue Genera und Species von Hydrozoen und anderen Evertebraten.

Arch. Naturgesch. 87: 248-265.

TREBILCOCK, R. E. 1928. Notes on New Zealand Hydroida. Proc. R. Soc. Vict. 41: 1-31.

VAN GEMERDEN-HOOGEVEEN, G. C. H. 1965. Hydroids of the Caribbean: Sertulariidae,

Plumulariidae and Aglaophentidae. Stud. Fauna Curagao 22: 1-87.

VERVOORT, W. 1941. Biological results of the Snellius Expedition. XI. The Hydroida of the

Snellius Expedition (Milleporidae and Stylasteridae excluded). Temminckia 6: 186-240.

VeERVOORT, W. 1959. The Hydroida of the tropical west coast of Africa. Atlantide Rep. 5:

211-325.

VERVOORT, W. 1968. Report on a collection of Hydroida from the Caribbean region, including

an annotated check-list of Caribbean hydroids. Zool. Verh., Leiden 92: 1-124.

WARREN, E. 1908. Ona collection of hydroids, mostly from the Natal coast. Ann. Natal Mus.

1: 269-355.

WEILL, R. 1937. Contribution a l’étude des Pteronemida (Hydrozoaires).—Le cnidome de

Cladocoryne floccosa Rotsch et son interprétation (d’apres les documents laissés par feu

Maurice Philbert). Bull. Inst. oceanogr. Monaco 719: 1-11.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in paren-

theses, volume number, part number (only if independently paged) in parentheses,

pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.-H. 1948. Données sur la résistaince et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FISCHER, P.-H., DUVAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 1960b. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

“

‘

Ft ay

N. A. H. Millard & J. Bouillon

A COLLECTION OF HYDROIDS

FROM MOCAMBIQUE, EAST AFRICA

: a ree

_ VOLUME 65 PART 2 MAY 1974

ANNALS

"OF THE SOUTH AFRICAN

MUSEUM

CAPE TOWN

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 #£2Band

May 1974 Mei

Part 2 Deel

REVISED LIST OF THE PRESERVED MATERIAL OF

THE EXTINCT CAPE COLONY QUAGGA,

EQUUS OQUAGGA QUAGGA (GMELIN)

R. E. RAU

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 20, 3, 5, 7-8), 3(1-2, 5, t.—p.i.), 51-2, 5, 7-9),

6(1, t.—p.i.), 711-3), 8, 911-2), 10(1),

111255, 7, t:—ps:);, 242); 27, 3@—3)533

Price of this part/Prys van hierdie deel

R6,00

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 39 9

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

REVISED LIST OF THE PRESERVED MATERIAL OF THE EXTINCT

CAPE COLONY QUAGGA, EQUUS QUAGGA QUAGGA (GMELIN)

R. E. Rau

South African Museum, Cape Town

(With 25 figures)

[MS accepted 12 June 1973]

CONTENTS

PAGE

Introduction . ‘ : ; 4 : ae?

List of institutions housing quagga material . . 48

Lost material . y 3 : ' ; SO

Description of preserved material : : Sm fe

Acknowledgements . : 2 : : wT 85

References : s : : ; : See 5)

INTRODUCTION

Much has been written about the reason for the extinction of the quagga,

Equus quagga quagga (Gmelin 1788). It is usually attributed to ‘ruthless’ hunting

and even ‘planned extermination’ by the colonists of the bigger herbivorous

animals, as these were considered to be competing against their livestock for

grazing. The quagga, which existed in large numbers, is also said to have been

hunted for its flesh, which the farmers fed to their labourers, and its skin, which

was used for grainbags and leather. Because of the great demand in the hide

trade for skins of quagga, zebra and other game, the farmers of the Orange

Free State in the eighteen-sixties organized hunting expeditions to collect them.

Wagon-loads of prepared skins were driven to the coast for sale (Bryden 1889:

401).

Eloff (1966) sees the reason for the extinction of the quagga in ‘an ever

greater pauperised gene pool relative to further migration southward, with

decreasing adaptability to changed environmental conditions’.

The quagga had a comparatively small range, which is a known factor in

the elimination of endemics. In addition, its range was restricted by farm fences

(Jackson 1920), which separated groups and individuals from one another.

While the above factors may all have contributed to the disappearance of

the quagga, there are another two which have not been mentioned by previous

authors.

South Africa was for a long time known as a ‘hunting paradise’. Many

books, e.g. Portraits of the game and wild animals of southern Africa (Harris

1840), give evidence of the senseless mass killing by those privileged to journey

to the Cape of Good Hope to satisfy their hunting urges. The quagga obviously

experienced this onslaught.

Ann. S. Afr. Mus. 65 (2), 1974: 41-87, 25 figs.

42 ANNALS OF THE SOUTH AFRICAN MUSEUM

The second factor was the years of drought, experienced after 1876, which

could have been the final coup de grace for the species. In his Manna in the desert,

Alfred de Jager Jackson (1920) recalls his younger days on a Great Karroo

farm near Nelspoort (the area from which the Cape Town quagga foal origi-

nated). He writes (p. 73): “The last years of my life on the Karroo farm were

clouded and sad. I would not care to live them again. As I have said, the good

seasons ended in 1876. The year following was a time of dread disease and

drought. The carcasses of thousands of dead animals lay rotting over the land.’

It has however also been stated that the species vanished without this

fact having been noticed by zoologists. In Afrikaans the word ‘Kwagga’ was,

and still is, used for the true quagga and for both plains zebra and the mountain

zebra. Outside the Cape Colony ‘Quagga’, in its various spellings, implied

Equus quagga quagga only. The first indigenous equids which the white settlers

in South Africa became acquainted with were the true quagga and the mountain

zebra. They referred to these as ‘Kwagga’, ‘Wilde-esel’ or ‘Wildeperd’,

‘“Gestreepte-esel’ or ‘sebra’ (Scholtz 1941). With their advance in a northerly

direction, the colonists met with Burchell’s zebra, which they called the ‘Bont-

kwagga’. Unfortunately in common usage in South Africa the ‘Bont’ is often

dropped and ‘Kwagga’ was and still is “good enough’ to describe any striped

equid.

It appears to be largely this confusion which led to the accidental disappear-

ance of the quagga. Had the true situation been realized, when the quagga

became rare in the middle of last century, efforts might have been made to

protect the species in its native country and to breed from those in European

zoos. (It should be noted that as early as 1822, when the need to protect the

bontebok, Damaliscus dorcas dorcas (Pallas), was realized, this antelope was

excluded from otherwise unspecified hunting licences, issued by the Colonial

Office in Cape Town.)

Although it cannot be determined when the last indigenous quaggas

disappeared, it is generally accepted that the female specimen which died at

Amsterdam Zoo on 12 August 1883 outlived those in South Africa. A letter

dated 6 November 1856 from A. Dale, Beaufort West District, to the director

of the South African Museum states that ‘Quaggas seem to be rather scarce’,

yet Mr Bols, the Belgian Consul at Port Elizabeth, still obtained ‘several’

quaggas in 1870; these were sent to Antwerp Zoo.

Several authors have discussed the striping and coloration of the quagga.

Old descriptions and illustrations which were made from living specimens

refer to animals with black or dark stripes on a brown or fawn base, while the

legs and ventral surface are given as white or whitish. Later authors declare

that these early descriptions were incorrect, due to carelessness, incompetence

and inaccurate observation, and describe the quagga as dark brown, chestnut

or fawn coloured with white or whitish stripes, restricted to the antero-dorsal

region of the animal, while other zebra forms are described as being light with

black-brown stripes. If the quagga is indeed a light-striped animal, how can the

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 43

presence of dark stripes or stripe fragments, posterior to the ‘light-striped’ portion

and above the hooves, be explained ?

With this in mind all quagga skin material in Europe, except for the foetus

in Stockholm, the ‘Elgin head’ together with a mounted skin in Edinburgh

and the material possibly existing in Russia, was examined by the author in 1971.

In common practice, the part which covers the smaller percentage of a

bi-coloured surface is called the ‘pattern’. The quagga has retained light inter-

spaces on the anterior portion of the body. These interspaces are, in contrast

to other members of the plains (or Burchell’s) zebra group, only half, or less

than half, as wide as the stripes. As these light regions cover a lesser proportion

of the animal than the dark, it has in publications of this century been described

as a light-striped animal.

As an adaptation to open country, the plains zebra group, including the

quagga, exhibits progressive reduction in striping and contrast between stripes

and interspaces from the north to the south of its range, i.e. the white darkens

towards brown, the black lightens to a brown. This process, which commences

from the hooves and from the buttocks, culminates in the disappearance of

stripes.

The quagga represents the extreme limit of this trend. The unstriped

postero-dorsal region of the animal is of a brown tone, which is intermediate

between the light and dark colours of the clearly striped antero-dorsal region.

The difference in coloration between northern and southern sub-species

of plains zebra appears to have been achieved by changing the pigment distribu-

tion. It seems therefore that in the quagga the proportion of light and dark

pigments is the same as in other members of the plains zebra group.

Examination of single hairs from various portions of the quagga reveals

that the areas between striped and unstriped regions are occupied by multi-

coloured hairs. Each hair shows light and dark transverse sections, similar to

the colour pattern of individual porcupine quills. Hairs from faint stripes have a

higher proportion of dark sections than hairs from the interspaces. The inter-

spaces, which show progressive darkening towards the posterior region of the

animal, may be distinguished, even in the regions with faint stripes or stripe

fragments, by their lighter colour. Hairs from other parts of the body are uniform

in colour and are light, dark or of intermediate tone, according to their position.

This type of pigmentation is particularly obvious in the Berlin, Vienna and

Amsterdam specimens.

Parallel with this reduction of contrast and striping is the occurrence

and progressive intensification of shadow-stripes (darkish streaks within the

interspaces). In southern forms of the plains zebra this has reached the stage

where, on the buttocks, it has become difficult to distinguish stripes from shadow-

stripes.

In the quagga this has been carried further. The region anterior to the

unstriped area is covered with numerous narrow, dark stripes or stripe frag-

ments, about twice the number as in the similar region of a plains zebra. Some

44 ANNALS OF THE SOUTH AFRICAN MUSEUM

of these narrow bands are obviously stripes while others are shadow-stripes.

This can be seen clearly in the Basle and Paris specimens.

The next stage in this process appears to be the fusion of shadow-stripes

with the true stripes, which produces the wide stripes characteristic of the

quagga. Often fragments of interspaces are retained, forming light dots or

streaks within the stripes. In regions where there is little fusion between frag-

ments of stripes and shadow-stripes, a dappling effect is often produced. This

has presumably led to illustrations like the one by Ridinger.

Whether the quagga represents a species or subspecies, and whether

differentiation by colour characteristics can be applied, have been much dis-

cussed. Hilzheimer (1912) considered that the narrow interspaces (hell gestreift)

were diagnostic. He therefore accepted the Wiesbaden specimen as a true

quagga, although he could not fit this specimen into the stripe-reduction

sequence suggested by Ridgeway (1909).

At first sight, the narrow interspaces and very reduced body striping appear

to be characteristic of the quagga, as is the reduced striping on the face (between

eye and corner of mouth) and broad dorsal median stripe, together with its

indicated or continuous flanking band on each side. However, all these characters

are variable and are independent of one another. On the basis of only one

of these four colour characteristics there is no sharp division between Equus q.

burchelli and Equus q. quagga. The preserved quagga specimens continue the

gradual change in colour and in marking shown by the north to south colour

variants of the plains zebra group. The advanced example would have no facial

striping, body stripes terminating at the shoulder, interspaces one-third or less

the width of the stripes, and a broad dorsal median stripe with a continuous

flanking band on each side.

Because of missing markings below the fork of the shoulder-stripe and

below the last body stripe, Hilzheimer (1912) erected the subspecies ‘paucistria-

tus’ of Equus burchelli, for two of the four Mainz specimens. (He considered

that the ‘lack’ of the ventral median stripe in the type might be characteristic,

but traces of this stripe can still be seen in the specimen. The Darmstadt and

Munich specimens also show this and there is no doubt that the ‘absence’ of

the ventral median stripe must be attributed to taxidermy. Furthermore the

second specimen of ‘Equus burchelli paucistriatus’ in Mainz possesses a ventral

median stripe.)

If Equus quagga quagga can be identified by the presence of at least two

of the four characteristics, then the long-haired female ‘Equus burchelli pauci-

striatus’ at Mainz is a true quagga, since it possesses narrow interspaces, faintly

striped face and broad dorsal median stripe, flanked with almost continuous

bands. Similarly the specimens at Tring and Vienna must be identified as

Equus quagga quagga. The former has a faintly striped face and broad dorsal

median stripe, flanked with interrupted light bands. The latter, in spite of mark-

ing below the shoulder-stripe fork and last body stripe has narrow interspaces

and a broad dorsal median stripe, flanked with interrupted light bands. These

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 45

specimens represent the variation of the quagga closest to Equus q. burchelli,

while the type of ‘Equus burchelli paucistriatus’ represents the variation of

Equus q. burchelli closest to Equus q. quagga. Thus the museum at Mainz has

three true quaggas, as suggested by Schwarz (1912), and also, of the preserved

specimens of Equus q. burchelli, the closest to the true quagga.

To summarize, the colour and markings of the preserved quagga specimens

do not support the identification of the quagga as a separate species, as suggested

by the following authors, some of whom have studied skulls and skeletons as

well: Hilzheimer (1912), Cabrera (1936), Allen (1939, 1945), Cooke (1943),

Lundholm (1951), Roberts (1951), Ellerman et al. (1953), Meester (1964),

Ansell (1967).

The question has been raised as to whether or not the colour and pattern

variation within the preserved quagga skins represents geographical forms.

As only a few skins have precise locality data, this has remained unanswered.

However, Antonius (1931) concluded that quaggas with both few and many

stripes might have occurred throughout the distribution area. This seems to be

confirmed by the following extract from a letter dated 11 June 1857 (1858)

to the director of the South African Museum from A. Dale of Kampherskraal

at “Nell’s Poort’ (Beaufort West District) announcing the arrival of the Cape

Town foal: ‘I believe there are two kinds, the stripes of the one kind being more

indistinct and of much paler colour than those of the other: this is a specimen

of the dark striped kind; the one which the Governor possesses and which was

reared at Nell’s Poort is of the other description.’ (Dale’s differentiation might

be due to the much darker appearance of a less striped animal compared with

a more extensively striped individual.) The colour variation within the pre-

served skins, where not due to fading or individual variation, could however

also be the result of seasonal changes.

An interesting point arises in connection with the Governor Sir George

Grey’s specimen. On 4 September 1858 he donated a quagga to London Zoo,

Regent’s Park. This male, of which a high-quality drawing by H. Weir appeared

in Illustrated London News, volume 33, 6 November 1858, had to be killed

because of self-inflicted injury on 10 June 1864. The mounted skin and skeleton

at the British Museum (Natural History), London, which were entered in the

catalogue on 2 July 1864, were regarded as those of Grey’s animal. However,

I have examined the skin and have found it to be that of a female, while the

skeleton is that of a male, if the presence of large canines can be accepted as a

means of identification. The shattered left metacarpal, as Dr A. W. Gentry

of the British Museum pointed out to me, seems to indicate that the skeleton at

least is of Grey’s animal. It is feasible that when Grey’s animal became available

in 1864 the British Museum decided to buy a more characteristic and already

stuffed skin from the Zoological Society’s Museum, which was most likely that

of the first London Zoo specimen which had died in 1834 (Shortridge 1934)

and to obtain or keep only the skeleton of Grey’s animal.

In trying to identify the illustration of Grey’s animal with one of the stuffed

46 ANNALS OF THE SOUTH AFRICAN MUSEUM

skins of which no data are available, the following became evident. Grey’s

animal was peculiar in having fragments of transverse stripes on the rump,

resembling the gridiron pattern of Equus zebra. This atypical marking is found

only in the Wiesbaden specimen, a male, and it is considered that this specimen

is the skin of Grey’s animal. Furthermore, the Wiesbaden specimen apparently

does not contain a skull or footbones (a practice unusual in taxidermy at that

time), and the skin is repaired in the area of the left metacarpal, the region of the

self-inflicted injury. The Wiesbaden specimen was bought from the dealer

Frank of Amsterdam in 1865. It is known that Frank dealt in quagga material

with the dealer Edw. Gerrard of London (Tring specimen), but no records of

transactions regarding Grey’s dead animal, nor of its disposition by the London

Zoo, could be traced.

Although the Governor did have a private menagerie at Cape Town

(Anon 1858), there are no records of any other quaggas in it. If the above

assumption is correct and the Governor’s quagga referred to by Dale was the

male donated to London Zoo, i.e. the Wiesbaden specimen, then there is

confirmation that the two extreme forms of quagga, as exemplified by the Cape

Town and Wiesbaden specimens, occurred in the same locality.

Mention should also be made of the second London Zoo quagga (1851-72),

a female, which is the only quagga ever photographed alive in Europe. According

to previous authors the skin of this animal was not preserved, as it was bad

(Ridgeway 1909). However, comparison of the photographs with the stuffed

specimen in the Royal Scottish Museum, Edinburgh, reveals that the two are

one and the same specimen. It appears, therefore, that the skin was not dis-

carded, but was sold to the Museum in 1879 by Edw. Gerrard. The skeleton of

this specimen has recently been discovered in the collection of the Peabody

Museum of Natural History, Yale University (Willoughby 1966).

The locality given for the Leiden quagga—Steenbergen, Cape Colony—is

incorrect. The position of the only ‘Steenbergen’ ever recorded is in the Cape

Peninsula, well outside the known distributional range of the quagga. This

mistake apparently stems from one of the two labels included with this specimen.

The two labels give different data: (i) ‘obtained alive in 1826’; (ii) ‘15th June,

1827—Steenbergen’. As can be seen below, neither of the labels is likely to

apply to this specimen. It is possible that (1) refers to the female specimen dis-

cussed below, while (11) probably refers to a male mountain zebra specimen

(Equus zebra) sent to Leiden in conjunction with a quagga. This mountain zebra

could well have been collected at Steenbergen.

During the years 1827-33, the Cape Town based medical doctor and agent

for the Rijksmuseum van Natuurlijke Historie in Leiden, H. B. van Horstok,

sent three quagga specimens to Leiden. The first, a female, he obtained alive,

but unfortunately the animal died before shipment, so that only its skin and

skeleton were sent to Leiden, as announced in Van Horstok’s letter to the direc-

tor, dated 1 July 1827. The second specimen, a male, together with a male

mountain zebra, was shipped on the frigate Bellona and was received by the

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 47

director of the navy in Holland in July 1830. The third specimen, of indeter-

minate sex, which Van Horstok obtained from ‘Graaf Reinjet’ on 25 March

1831 was dispatched to Leiden on or about 20 May 1833, the date of the letter

which announced the dispatch.

As far as is known, the Rijksmuseum van Natuurlijke Historie handled

only these three quaggas of which there are today only one mounted stallion

and a skeleton at Leiden. However, the immature quagga skeleton at Berlin

is said to have been exchanged with the Rijksmuseum van Natuurlijke Historie

in 1833-8 (Opperman 1970).

The policy at the Rijksmuseum van Natuurlijke Historie at that time was

to keep a perfect male, female and juvenile specimen of every species and to

exchange or sell any additional or imperfect specimens. (Van Bruggen, personal

communication.)

On 3 August 1830 a quagga (skin and male skull) was dispatched from

Leiden to the Senckenberg Museum in Frankfurt, where it was received in 1831.

Renshaw (1904) suggested the possibility that the Turin specimen may have

been collected by Van Horstok. This female specimen, which was bought in

1827 by the Turin museum from the dealer S. Leadbeater of London, is damaged

and has a sewn cut on the right side of the dorsal median stripe. These facts,

when considered in the light of the policy of the Ryksmuseum van Natuurlijke

Historie, appear to suggest that the Turin specimen might be the first (female)

Leiden quagga. It is however difficult to imagine that the specimen which

arrived in Holland probably late in 1827 could have been mounted, sold to

Leadbeater and from there to Turin, all within that year.

Unfortunately no certainty about the Leiden quaggas could be reached.

The history of the Tring specimen must be mentioned. On 2 April 1842

Lord Derby requested a live male quagga from his Cape Town agent, the

Reverend John Fry (xerox copy of letter at the South African Museum, Cape

Town). Lord Derby had at his Knowsley menagerie a female quagga, which

might have been obtained from Fry. After Derby’s death his animals were sold

in 1851. It appears that the male quagga had died previously. The female was

sold to Amsterdam Zoo, and was stuffed when it died there in 1853. After the

last living quagga had died in 1883 and been mounted, the Amsterdam Museum _

may have sold its earlier specimen to the dealer Frank of Amsterdam, who is

known to have bought a mounted quagga as a ‘duplicate from a continental

museum’ (Renshaw 1904). Frank sold this specimen to Edw. Gerrard of London,

who re-mounted it and sold it in 1889 to Lord Rothschild of Tring. That the

Tring specimen might in fact be the Knowsley female seems proven by the

specimen’s old label ‘From E. Gerrard, Jun., 61 College Place, Camden Town,

London—Quagga Equus quagga Linn. Gray Knowsley Menagerie—South

Africa’. Supporting this is the resemblance between the Tring specimen and the

animal on the right in Hawkins’s drawing (Gray 1850) of the live Knowsley

quaggas. It is obvious that the artist did not make a completely correct picture,

nevertheless characteristics like the big light patch in one of the last of the

48 ANNALS OF THE SOUTH AFRICAN MUSEUM

body stripes on the left and the dorsal forking of some body stripes are found

in both.

Although there have been attempts to breed quagga in captivity there are

apparently no records of this. However there are records of the birth of quagga

hybrids (Wagner 1835; Renshaw 1904, 1935) with either a male or female

quagga parent. In one instance, the female horse-quagga hybrid proved to be

fertile and was crossed with an arab stallion, the resulting offspring resembling

the quagga in its mane and body striping (Wagner 1835).

None of these hybrids is preserved, but according to descriptions they

differed from hybrids of other zebra forms. These have clear, zebra-like leg

striping while the quagga hybrids have few leg stripes, mainly at the ‘knees’

and ‘hocks’, and striping on the neck and shoulder. This leg striping in the

hybrids, which should not be confused with faint markings just above the

hooves, traceable in most preserved quagga skins, is interesting since the quagga

parent in each case must have had plain coloured legs.

The stuffed equid foal in the Museum and Art Gallery at Doncaster,

Yorkshire, which has not been previously recorded in the literature, exhibits

the characteristics of quagga hybrids. It appears to be the only quagga hybrid

skin still in existence. While the label on the outside of the original showcase

reads: ‘Zebra-foal born at Owston, Ist April 1830’, there is an old label attached

to the specimen, naming it as a ‘cross between a male ass and a female quagga’.

This foal in its case was donated to the museum by the Davies-Cooke family

in 1923. Phillip Davies-Cooke brought a ‘zebra’ from the Cape in the early

nineteenth century. This animal was in all probability the mother of the foal

which is now in the Doncaster Museum. The skeleton which Davies-Cooke

donated to the Yorkshire Philosophical Society in 1841 is most probably that of

the mother.

Two further unusual equid hybrids, both with the same parents, are pre-

served in the Tring Museum. They were born in Lord Rothschild’s menagerie

at the end of the nineteenth century and are the result of a cross between a male

horse and a female Burchell’s zebra. Their striping is similar to that of the foal

at Doncaster, but is more intensive and extends further posteriorly. The larger

of the two shows quite strong but narrow dark stripes on the buttocks. It seems

quite likely that their mother was in fact a specimen of Equus quagga burchelli.

LIST OF INSTITUTIONS HOUSING QUAGGA MATERIAL

Some inaccurate statements concerning the preserved quagga material are

to be found in the literature, including the lists by Renshaw (1904), Ridgeway

(1909), Hilzheimer (1912), Harper (1945) and Rzasnicki (1949); these errors

are corrected in the following list. The osteological material not belonging to

skins is listed without guarantee of correct identification and is based only on

information given by the institutions concerned or previous authors.

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA

Place Sex

Amsterdam Q

Bamberg oO)

Basle g

Berlin Q

juy.

Bristol* 3

Cape Town 92 foal

Darmstadt Q

Edinburgh iy

Frankfurt 3

Kazan

Leiden 3

London 2

Mainz 3

foal

Milan °imm.

Munich 2

New Haven 2

Skin

head

Com-

plete

skele-

ton

| ar |

* See footnote on page 57.

Loose

Skull bones

+1 ++

Institution

Zodlogisch Museum

Naturkunde Museum

Naturhistorisches

Museum

Museum ftir Natur-

kunde an der

Address

Plantage Middenlaan 53

Amsterdam-C

Netherlands

Fleischstrasse 2

D-86 Bamberg

W. Germany

Augustinergasse 2

CH-4051 Basle

Switzerland

Invalidenstrasse 43

X-104 Berlin

Humboldt Universitat Germany

City Museum,

Department of

Natural History

South African

Museum

Hessisches Landes-

museum, Zoologische

Abteilung

Royal Scottish

Museum

Natur Museum

Senckenberg

Zoological Museum

Rijksmuseum van

Natuurlijke Historie

British Museum

(Natural History)

University College,

Department of

Zoology

Naturhistorisches

Museum

Museo Civico di

Storia Naturale

Zoologische Samm-

lung des Bayerischen

Staates

Peabody Museum

of Natural History,

Yale University

Queen’s Road

Bristol BS8 1RL

England

Queen Victoria Street

FOF (Box 61,

Cape Town,

South Africa

Friedensplatz 1

D-61 Darmstadt

W. Germany

Chambers Street

Edinburgh EH1-1JT

Scotland

Senckenberg-Anlage 25

D-6 Frankfurt/Main 1

W. Germany

University

Kazan, USSR

Raamsteeg 2

Leiden, Netherlands

Cromwell Road

London SW7 5BD

England

Gower Street

London WCIE 6BT

England

Reichklarastrasse 1

D-65 Mainz

W. Germany

Corso Venezia 55

I-20121 Milan

Italy

Schloss Nymphenburg

Nordfliigel

D-8 Munich 19

W. Germany

New Haven

Conn. 06520, U.S.A.

50 ANNALS OF THE SOUTH AFRICAN MUSEUM

Com-

plete Loose

Place Sex Skin skele- Skull bones Institution Address

ton

Paris 3 ae + _ — Museum National 55 rue de Buffon

d’Histoire Naturelle, Paris—S5Se

Laboratoire de France

Mammalogie

Philadelphia 4d “ 7 The Academy of Nineteenth and the

Natural Sciences Parkway

Philadelphia

Penn. 19103, U.S.A.

Pretoria Q = = Transvaal Museum ‘Paul Kruger Street

PO Box 413

Pretoria, Transvaal

South Africa

Stockholm foetus + — Naturhistoriska Roslagsvagen 124

Riksmuseet S-104 05 Stockholm 50

Sweden

Stuttgart 2 = = Staatliches Schloss Rosenstein

Museum fiir D-7 Stuttgart 1

Naturkunde W. Germany

Turin Q 4. — Museo e Instituto Via Gioletti 34

de Zoologia I— 10123 Turin

Sistematica Italy

Tring 2 = = Zoological Museum Akeman Street

(affiliated to the Tring, Herts.

British Museum England

(Natural History))

Tubingen 3 — — Zoologisches D-74 Tubingen

Institut der W. Germany

Universitat

Vienna 2 =F — Naturhistorisches Burgring 7

Museum A-1014 Vienna 1

Austria

Wiesbaden 3 + — Stadtisches Museum, Rheinstrasse 10

-+ = present;

Naturwissen-

D-62 Wiesbaden

schaftliche Abteilung W. Germany

— = not present;

(++) = present in mount; t = incisivae only.

LOST MATERIAL

Rzasnicki (1949) expressed the fear that a post-war inventory of quagga

material might reveal many losses. The following specimens have indeed been

lost, in some cases as a direct consequence of World War II.

AMSTERDAM: A female quagga skeleton was mounted for the Amsterdam

museum; the whereabouts of the specimen are unknown (Tuijn 1966).

GRAHAMSTOWN: The quagga skull was destroyed by fire in 1942.

KOENIGSBERG: The mounted specimen was lost when ‘Waldhof’, a near-by castle

used for safe storage by the museum, was burnt down at the end of the war.

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 51

LONDON: The two male skulls, housed in the Royal College of Surgeons, were

destroyed during the bombing raids on that city.

MAINZ: The foal was partly destroyed by fire during World War II. Only the

head, front legs and rump with hind legs and tail have been salvaged.

Schwarz (1912) mentioned a male quagga skull as being in the collection

of the Naturhistorisches Museum. This skull no longer exists nor are there

any records of it.

MANCHESTER: Renshaw (1904) lists a quagga skeleton at the medical museum of

the University. This specimen no longer exists, nor are there any records

of its fate.

STUTTGART: Although Hilzheimer (1912) has described and figured quagga foot

and leg bones housed in Stuttgart, these bones no longer exist nor is there

any record of their disposition. The bones belonged either to the skin sent

from Cape Town in 1827 by Von Ludwig to Tubingen or to the Amsterdam

specimen.

YORK: A quagga skeleton was donated to the Yorkshire Philosophical Society

(Annual Report 1841) by Ph. Davies-Cooke. The specimen no longer

exists, nor are there any further records.

DESCRIPTION OF PRESERVED MATERIAL

EXPLANATION OF TERMS

Face: area between eye and corner of mouth

Interspace: light portion of colour pattern

Stripe: dark portion of colour pattern

Head-body: following contour of dorsal mid-line, except for portion between

upper lip and posterior margin of nostrils and middle of rump to extreme

posterior margin of buttocks, where the ruler remained straight. From

anterior end of muzzle and posterior end of buttock a line to meet ruler

in a 90° angle was imagined

Tail: from where it leaves body to flesh tip, leaving out the brush

Ear: along mid-line of outer surface, from head to tip

Hindfoot: along lateral side from ‘sole’ of hoof to middle of fetlock, from there

to middle of heel

Shoulder height: vertically from base to highest point of withers

All measurements were taken by the author except for the specimens at

Edinburgh, Doncaster and Stockholm.

Bones belonging to skins: not listed separately.

Bones belonging to skins but at different institutions: listed separately.

The references cited exclude the numerous illustrations, with or without

short notes, when these contribute no specific information.

352 ANNALS OF THE SOUTH AFRICAN MUSEUM

The type of Equus burchelli paucistriatus (Hilzheimer 1912) is included as

the coloration of the specimen is intermediate between that of Equus quagga

burchelli and Equus quagga quagga. It was considered synonymous with the true

quagga by Allen (1939). The foal at Doncaster is included, as it appears to be

the only quagga hybrid preserved.

AMSTERDAM— mounted skin and skull

Fic. 1. Amsterdam quagga.

Catalogue number: ZMA 522

Sex: female

Locality: ——

Date of acquisition: 1883

Remarks on acquisition: animal lived at Amsterdam Zoo, 9 May 1867-12 August

1883

History of mount: original mount by Inspector Kerz (Hilzheimer 1912: 91)

Description of striking features: face fairly striped; light flanking bands of dorsal

median stripe not interrupted; hair very short, transversely multicoloured

in some regions; dark portions of animal umber, not the usual chestnut

colour

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 53

Measurements: head—body: 2,28 m

tail: 0,46 m

ear: 0,16 m

hindfoot: 0,46 m

shoulder height: 1,20 m

State of preservation: very good

Further material of same individual: skull in collection. It appears from Hilz-

heimer (1912: 98) that the foot bones of this specimen were at Stuttgart

but they are no longer there, nor is there any record of them

Remarks: not exhibited

References: Lydekker (1904); Renshaw (1904); Ridgeway (1909); Hilzheimer

(1912); Griffini (1913); Antonius (1931); Van Bruggen (1959)

BAMBERG— mounted skin

re > . i \

oN _ ri P

a. asl

Fic. 2. Bamberg quagga.

Catalogue number: 236 (mammal catalogue)

Sex: female?

54 ANNALS OF THE SOUTH AFRICAN MUSEUM

Locality: ——

Date of acquisition: 9 November 1858

Remarks on acquisition: bought as stuffed specimen from Dr F. Krauss, Stuttgart

(Antonius 1931)

History of mount: re-mounted in 1969 by dermo-sculptor Kaestner, Berlin

Description of striking features: face unstriped; light flanking bands of dorsal

median stripe not interrupted

Measurements: head—body: 12935) im

tail: 0,410 m

ear: 0,165 m

hindfoot: 0,490 m

shoulder height 1,100 m

State of preservation: good

Further material of same individual: ——

Remarks: exhibited open; alien matter used for improvement

References: Schwarz (1912); Antonius (1931)

BASLE— mounted skin, skull and foot bones

“ Deere

Fic. 3. Basle quagga.

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 28)

Catalogue number: 897

Sex: female

Locality: received from Shiloh/Whittlesea, Eastern Cape Province

Date of acquisition: 1862/63

Remarks on acquisition: donated by missionary S. Gysin at Shiloh

History of mount: re-mounted in 1929 by dermo-sculptor G. Ruprecht

Description of striking features: face faintly striped near eye; light flanking

bands of dorsal median stripe several times interrupted in anterior portion

Measurements: head—body: 2,14 m

tail: 0,38 m

ear: 0,15 m

hindfoot: 0,47 m

shoulder height: 1,20 m

State of preservation: good

Further material of same individual: skull (No 2860) and skeletal parts (No 8099,

10304) in collection

Remarks: exhibited

References: Ridgeway (1909); Roux (1910); Griffini (1913); Antonius (1931)

BERLIN—mounted skin and skull

Fic. 4. Berlin quagga.

56 ANNALS OF THE SOUTH AFRICAN MUSEUM

Catalogue number: 4832 (old number: A1133)

Sex: female

Locality: ——

Date of acquisition: 1867

Remarks on acquisition: animal lived at Berlin Zoo, 1863-7

History of mount: original mount (stuffed with straw)

Description of striking features: face clearly striped; light flanking bands of

dorsal median stripe not interrupted; animal very dark; hairs of faint

striped regions with light and dark transverse sections

Measurements: head—body: 2,20 m

tail: 0,35 m

ear: 0,15 m

hindfoot: 0,47 m

shoulder height: 1,11 m

State of preservation: good, probably not faded

Further material of same individual: skull in collection

Remarks: not exhibited

References: Renshaw (1904); Ridgeway (1909); Hilzheimer (1912); Antonius

(1931, 1951); Opperman (1970); Liversidge & ffolliott (1971)

BERLIN— skeleton

Catalogue number: 38954 (old number: An 8954)

Sex: immature

Locality: ——

Date of acquisition: 1833-8

Remarks on acquisition: from anatomical collection J. Miiller, exchanged with

Ryksmuseum van Natuurlijke Historie, Leiden

Further material of same individual: ——

References: Renshaw (1904); Ridgeway (1909); Hilzheimer (1912); Antonius

(1931); Opperman (1970)

BERLIN— skull

Catalogue number: 23707 (old number An 1407)

Sex: male

Locality: ——

Date of acquisition: 1833-8

Remarks on acquisition: from anatomical collection J. Miiller

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA ay

Further material of same individual: ——

References: Renshaw (1904); Ridgeway (1909); Hilzheimer (1912); Antonius

(1931, 1951); Opperman (1970)

BRISTOL — skull*

Catalogue number: Aa 3294

Sex: male

Locality: ——

Date of acquisition: probably 1927

Remarks on acquisition: registered 1927

State of preservation: right PM, missing

Further material of same individual; ——

CAPE TOWN—mounted skin, skull and footbones

Fic. 5. Cape Town quagga foal.

Catalogue number: SAM 35575

Sex: female, foal

Locality: Nelspoort, Beaufort West District, Cape Province

* In January 1974, while this paper was in press, Dr A. W. Gentry identified this skull as

that of Equus caballus.

58 ANNALS OF THE SOUTH AFRICAN MUSEUM

Date of acquisition: 1857-60

Remarks on acquisition: donated by A. Dale of Campherskraal, Nelspoort;

animal lived one week after capture

History of mount: stuffed with hemp and clay at South African Museum;

re-mounted by dermo-sculptor R. Rau 1969/70

Description of striking features: specimen with long, woolly fur (25-30 mm long);

face faintly striped; light flanking bands of dorsal median stripe incon-

spicuous, not interrupted

Measurements: head-body: 1,240 m

tail: 0,280 m

Cale: 0,123 m

hindfoot: 0,390 m

shoulder height: 0,770 m

State of preservation: good, partly moth-eaten

Further material of same individual: skull, footbones and fleshy parts of skin,

all removed during re-mounting, in collection

Remarks: exhibited; since 1962 in special fade-ban glass case

References: Renshaw (1904, 1909); Ridgeway (1909); Antonius (1931); Short-

ridge (1934)

DARMSTADT— mounted skin

Fic. 6. Darmstadt quagga.

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 59

Catalogue number: HLM, M719

Sex: female

Locality: ——

Date of acgisition: 1830

Remarks on acquisition: skin obtained from Stuttgart Museum (Scheer in Tit.

1959); may possibly be the skin received in 1827 from Baron C. F. H.

von Ludwig of Cape Town (Hilzheimer 1912)

History of mount: re-mounted in 1862 by Inspector Kerz (Hilzheimer 1912)

Description of striking features: face clearly striped; light flanking bands of

dorsal median stripe twice interrupted; ventral median stripe lost through

taxidermy, except for anterior portion

Measurements: head—body: 2,260 m

tail: 0,520 m

ear: 0,165 m

hindfoot: 0,500 m

shoulder height: 1,200 m

State of preservation: good

Further material of same individual: it is likely that the skull at Stuttgart belongs

to this animal

Remarks: exhibited

References: Hilzheimer (1912)

DONCASTER— mounted skin

(Hybrid Equus gq. quagga 2° & Equus asinus $)

FiG: 23

Catalogue number: 108.26

Sex: foal

Locality: born at Owsten, Yorkshire, | April 1830

Date of acquisition: 1923

Remarks on acquisition: foal died soon after birth; mounted by Hugh Reid,

Doncaster, for Phillip Davies-Cooke, whose family donated it to the

museum

History of mount: original mount

Description of striking features: faint striping traceable in most parts; several

stronger stripes at withers and legs below ‘knee’ and ‘hock’, tail with long

hair from the root, lower portion forming brush; mane and tail uniformly

dark

60 ANNALS OF THE SOUTH AFRICAN MUSEUM

Measurements: head—body: 1,170 m

tail: 0,280 m

Gan: 0,105 m

hindfoot: 0,400 m

shoulder height: 0,780 m

State of preservation: good, probably faded

Further material of same individual: ——

EDINBURGH— mounted skin

Fic. 7. Edinburgh quagga.

Catalogue number: 1879.35.1

Sex: female

Locality: ——

Date of acquisition: 1879

Remarks on acquisition: bought from Edw. Gerrard, London; skin of London

Zoo female, 15 March 1851 to 7 July 1872

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 61

History of mount: original mount

Description of striking features: face fairly striped; light flanking bands of dorsal

median stripe not interrupted

Measurements: head—body: 2,205 m

tail: 0,400 m

ealeuicite: 0,150 m

ear, right: 0,145 m

hindfcot: 0,430 m

shoulder height: 1,170 m

State of preservation: good

Further material oj same individual: skeleton at Peabody Museum of Natural

History, Yale University, New Haven, U.S.A.

Remarks: Specimen is second Royal Scottish Museum quagga; first specimen

(bought 1813, not 1818 as stated in the literature) no longer exists

References: Renshaw (1904); Ridgeway (1909); Antonius (1931)

EDINBURGH— mounted head

Fic. 8. Edinburgh quagga head (formerly Elgin).

62 ANNALS OF THE SOUTH AFRICAN MUSEUM

Catalogue number: 1970-67

Sex: ——

Locality: King William’s Town, Cape Province

Date of acquisition: 1970

Remarks on acquisition: 1861 presented by John Maclean of King William’s

Town to Elgin and Morayshire Society, Elgin; 1970 sold to Royal Scottish

Museum

History of mount: original mount, likely to have originated from complete

mount

Description of striking features: face unstriped—probably through fading

Measurements: head, upper lip

to between ears: 0,415 m

ear, left: 0,170 m

ear, right: 0,150 m

State of preservation: much faded

Further material of same individual: skull is obviously inside mount

References: Ridgeway (1905, 1909)

FRANKFURT— mounted skin and skull

Fic, 9. Frankfurt quagga.

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 63

Catalogue number: 19207

Sex: registered as male, skin has no sexual organs

Locality: ——

Date of acquisition: 1831

Remarks on acquisition: by exchange from Rijksmuseum van Natuurlijke

Historie, Leiden; probably one of the three quaggas received by the Rijks-

museum between 1827 and 1833 from their agent at Cape Town, Dr H. B.

van Horstok

History of mount: remounted at Frankfurt (Lotichius 1912)

Description of striking features: face clearly striped; light flanking bands of

dorsal median stripe, where preserved, only slightly interrupted

Measurements: head—body: 2,100 m

tail: 0,450 m

ear: 0,145 m

hindfoot: 0,450 m

shoulder height: 1,090 m

State of preservation: good, much patched

Further material of same individual: male skull in collection, occipital missing

Remarks: exhibited

References: Hilzheimer (1912); Lotichius (1912), Antonius (1931)

KAZAN—mounted skin

Catalogue number:

Sex:

Locality:

Date of acquisition: 1843

Remarks on acquisition: bought in Hamburg from a Mr Brandt by Professor

Eversman for the Zoological Museum of Kazan University.

History of mount: it appears that this specimen was purchased as an old mount

and remounted. In 1969 Mr Zaslavsky remounted specimen again

Description of striking features:

Measurements:

State of preservation:

Further material of same individual:

Remarks: the details given here were supplied to Dr V. Eisenmann, Paris, by

Professor V. A. Popov, in charge of the Zoological Museum of Kazan

University. No further information could be obtained directly from Kazan.

64 ANNALS OF THE SOUTH AFRICAN MUSEUM

LEIDEN— mounted skin and complete skeleton

Fic. 10. Leiden quagga.

Catalogue number: 18243 (old number: Cat. Jentink 1892 No. a)

Sex: male

Locality: Cape Colony

Date of acquisition: 1830-3

Remarks on acquisition: specimen is probably the second or third of the three

received by the Rijksmuseum van Natuurlijke Historie between 1827 and

1833 from their agent at Cape Town, Dr H. B. van Horstok

History of mount: original mount of straw. Specimen exhibited until 1913,

causing slight fading of left side

Description of striking features: face clearly striped; mane hair very short;

posterior neck stripes breaking up ventrally; light flanking bands of dorsal

median stripe twice slightly interrupted anteriorly; posterior striping

forming dappling; body stripes and dorsal median stripe with light streaks

and dots; faint dark transverse bands above hooves

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 65

Measurements: head—body: 2,000 m

tail: 0,430 m

ear, left: 0,180 m

ear, right: 0,165 m

hindfoot: 0,470 m

shoulder height: 1,100 m

State of preservation: very good

Further material of same individual: complete skeleton in collection

Remarks: not exhibited; housed in total darkness

References: Renshaw (1904); Ridgeway (1909); Antonius (1931); Van Bruggen

(1959)

LONDON, BRITISH MUSEUM (NATURAL HISTORY)—mounted skin

Fic. 11. London quagga.

Catalogue number: 1864.7.2.3. (old number: 1449a)

Sex: female

Locality: ——

Date of acquisition: 1864

Remarks on acquisition: it seems probable that this specimen is the first London

Zoo quagga, 5 November 1831-4 (Shortridge 1934; Antonius 1931;

Sclater 1901). In 1864 the British Museum bought a stuffed quagga from

66 ANNALS OF THE SOUTH AFRICAN MUSEUM

the museum of the Zoological Society and in the same year received at least

the skeleton of the third London Zoo quagga, 4 September 1858-10 June

1864 (Sclater 1901; Renshaw 1904)

History of mount: original mount of straw

Description of striking features: face faintly striped; light flanking bands of

dorsal median stripe twice interrupted on left

Measurements: head—body: 2,400 m

tail: 0,410 m

ear: 0,165 m

hindfoot, left: 0,480 m

hindfoot, right: 0,440 m

shoulder height: 1,180 m

State of preservation: poor, left ear off but present; skin cracked; crude repairs

Further material of same individual: the complete skeleton in the collection,

which was believed to belong to this skin, is that of a male and probably

that of the third London Zoo quagga whose skin is considered to be the

one at Wiesbaden Museum

Remarks: not exhibited

References: Lydekker (1904); Renshaw (1904); Ridgeway (1909); Flower

(1929); Antonius (1931)

LONDON, BRITISH MUSEUM (NATURAL HISTORY)—Sskeleton

Catalogue number: 1864.7.2.3. (old number: 1449a)

Sex: male

Locality: probably ‘Nell’s Poort’, Cape Province

Date of acquisition: 2 July 1864

Remarks on acquisition: live animal received at London Zoo 4 September 1858

State of preservation: left metacarpal shattered

Remarks; it is considered that this skeleton belongs to the mounted skin at

Wiesbaden, Germany

References: Sclater (1901); Lydekker (1904); Renshaw (1904); Ridgeway (1909)

LONDON, UNIVERSITY COLLEGE— Skeleton

Catalogue number:

Sex: female

Locality:

Date of acquisition:

Remarks on acquisition;

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 67

State of preservation: left hind limb and right scapula missing

Remarks: Dr A. W. Gentry, British Museum (Natural History), is of the opinion

that the specimen is Equus quagga quagga

MAINZ— mounted skin

Fic. 12. Mainz, male quagga on left, Burchell’s zebra (type of Equus quagga paucistriatus

Hilzheimer, 1912) on right.

Catalogue number: 1955/13

Sex: male

Locality: ——

Date of acquisition: 1840-50 (Hilzheimer 1912)

Remarks on acquisition: bought from dealer Rtihl, Wiesbaden

History of mount: original mount

Description of striking features: face unstriped; light flanking bands of dorsal

median stripe several times interrupted; shadow stripes on neck

Measurements: head—body: 2,020 m

tail: 0,365 m

ear: 0,150 m

hindfoot: 0,440 m

shoulder height: 1,130 m

68 ANNALS OF THE SOUTH AFRICAN MUSEUM

State of preservation: fair, skin cracked

Further material of same individual: ——

Remarks: exhibited

References: Hilzheimer (1912); Schwarz (1912); Antonius (1931)

MAINZ— mounted skin

Fic. 13. Mainz, female quagga.

Catalogue number: 1955/11

Sex: female

Locality: ——

Date of acquisition: 1840-50 (Hilzheimer 1912)

Remarks on acquisition: bought from dealer Riihl, Wiesbaden

History of mount: original mount

Description of striking features: hair long, woolly; face faintly striped; light

flanking bands of dorsal median stripe not interrupted, except for narrow

contact with Ist pair of body stripes

Measurements: head—body: 2,220 m

tail: 0,430 m

ear: 0,135 m

hindfoot: 0,430 m

shoulder height: 1,160 m

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 69

State of preservation: fair; skin cracked

Further material of same individual; ——

Remarks: exhibited

References: Hilzheimer (1912); Schwarz (1912); Antonius (193i)

MAINZ— mounted skin

Fic. 14. Mainz, rescued portions of burnt foal.

Catalogue number: 1955/14

Sex: foal

Locality; ——

70 ANNALS OF THE SOUTH AFRICAN MUSEUM

Date of acquisition: 1840-50 (Hilzheimer 1912)

Remarks on acquisition: bought from dealer Ruhl, Wiesbaden

History of mount: original mount

Description of striking features: face faintly striped; hair short

Measurements: head from upper lip

to between ears, straight: 0,245 m

tail: 0,190 m

hindfoot: 0,285 m

height at rump: 0,630 m

State of preservation: fair, partly destroyed by fire

Further material of same individual; skull inside mount (occipital burnt)

Remarks: not exhibited

References: Hilzheimer (1912); Schwarz (1912); Antonius (1931)

MAINZ—mounted skin

(Type of Equus burchelli paucistriatus)

Fic. 12

Catalogue number: 1955/12

Sex: female

Locality: ——

Date of acquisition: 1840-50 (Hilzheimer 1912)

Remarks on acquisition: bought from dealer Ritihl, Wiesbaden

History of mount: original mount

Description of striking features: face clearly striped; interspaces wide with

shadow stripes; light flanking bands of dorsal median stripe present from

lumbar region only; no markings below shoulder stripe fork and last

body stripe

Measurements: head—body: 2,300 m

tail: 0,420 m

ear: 0,165 m

hindfoot: 0,460 m

shoulder height: 1,320 m

State of preservation: fair, skin cracked

Further material of same individual: ——

Remarks; exhibited

References: Hilzheimer (1912); Schwarz (1912); Antonius (1931)

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 71

MILAN —mounted skin

Fic. 15. Milan quagga.

Catalogue number: not registered

Sex: female, immature

Locality: ——

Date of acquisition: before 1848 (Sordelli 1909)

Remarks on acquisition:

History of mount: original mount

Description of striking features: face unstriped; light flanking bands of dorsal

median stripe several times interrupied; coat long and woolly

Measurements: head—body: 1,820 m

tail: 0,370 m

ear: 0,175 m

hindfoot, left: 0,460 m

hindfoot, right: 0,435 m

shoulder height: 1,020 m

a2 ANNALS OF THE SOUTH AFRICAN MUSEUM

State of preservation: fair, much faded, partly moth-eaten

Further material of same individual: skull or part thereof inside mount, upper

and lower incisors visible

Remarks: exhibited

References: Sordelli (1909); Griffini (1913)

MUNICH— mounted skin and incisivae

Fic. 16. Munich quagga.

Catalogue number: AM 541

Sex: female

Locality: ——

Date of acquisition: 1834—5

Remarks on acquisition: bought from C. F. Ecklon, Hamburg, probably as

unmounted skin

History of mount: mounted or remounted during first decade of this century by

Inspector Kiisthardt

Description of striking features: face clearly striped; light flanking bands of

dorsal median stripe not interrupted; most of ventral median stripe lost

through taxidermy

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 73

Measurements: head—body: Peay ie)

tail: 0,42 m

ear: 0,16 m

hindfoot: 0,46 m

shoulder height: 1,17 m

State of preservation: much cracked; specimen very light, probably faded

Further material of same individual: upper and lower incisivae in collection

Remarks; not exhibited

References: Wagner (1835); Renshaw (1904); Ridgeway (1909); Hilzheimer

(1912); Antonius (1931, 1951)

MUNICH — skull

Catalogue number: AM 54la

Sex: female

Locality: ——

Date of acquisition: 1834-5

Remarks on acquisition: bought from C. F. Ecklon, Hamburg, together with skin

State of preservation: occipital missing

Further material of same individual; ——

References: Hilzheimer (1912)

MUNICH—upper and lower incisivae

Catalogue number: AM 561

Sex: male

Locality: ——

Remarks on acquisition: ——

Further material of same individual: ——

NEW HAVEN—Skeleton

Catalogue number: Osteology 1623 (490)

Sex: female

Locality: given as ‘Syria (?)’

Date of acquisition: 1873

Remarks on acquisition: bought by O. C. Marsh from Edw. Gerrard, London;

of same animal as mounted skin at Edinburgh

State of preservation; some molars missing

74 ANNALS OF THE SOUTH AFRICAN MUSEUM

Remarks: not articulated, cleaned 1923

References: Willoughby (1966)

PARIS— mounted skin and skeleton

‘

Fic. 17. Paris quagga.

Catalogue number: not registered

Sex: male

Locality: Cape Colony

Date of acquisition: probably 1798

Remarks on acquisition: animal arrived at the Gardens of the King at Versailles

in 1784 from the Cape Colony; was transferred to Paris (Jardin des Plantes)

in 1793 or 1794 (Dorst 1952)

History of mount: original mount, mainly of wood

Description of striking features: face clearly striped; mane short; sudden change

from dorsal colour to leg colour at elbow; light flanking bands of dorsal

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 15

median stripe not interrupted; tail brush very long (0,49 m beyond tip of

tail), most probably not original as fixing on to tail can be seen; skin of

chestnuts on front legs replaced by white-haired skin

Measurements: head—body: 1,93 m

tail: 0,29 m

Cate: Osl5om

hindfoot: 0,43 m

shoulder height: 1,18 m

State of preservation: good

Further material of same individual: there is an articulated skeleton (No A544)

labelled “Quagga’ in the Department of Comparative Anatomy, Paris

Museum. It is not certain whether it belongs to the skin.

Remarks; not exhibited, stored in glass case

References: Desmarest (1820); Renshaw (1904); Trouessart (1906); Pocock

(1907); Ridgeway (1909); Griffini (1913); Antonius (1931); Dorst (1952)

PHILADELPHIA — Skeleton

Catalogue number: ANS 6317

Sex: male

Locality: ——

Date of acquisition: 1898

Remarks on acquisition: donated by Professor E. D. Cope

References; Renshaw (1904)

PRETORIA — skull

Catalogue number: TM 10161

Sex. female

Locality: ——

Date of acquisition: ‘long ago’

Remarks on acquisition: identified in old collection by Lundholm

State of preservation: complete

References: Lundholm (1951)

76 ANNALS OF THE SOUTH AFRICAN MUSEUM

STOCKHOLM— mounted skin

Fic. 18. Stockholm quagga foetus.

Catalogue number: Mam Ex. 14

Sex. foetus

Locality: Cape Coiony

Date of acquisition: 1775

Remarks on acquisition: brought to Sweden by A. Sparrman

History of mount: original mount of straw; repairs with two hand-sized white

leather patches might be of later date

Description of striking features: face clearly striped; light flanking bands of

dorsal median stripe not interrupted; interspaces fairly wide

Measurements: head-body: 1,020 m

tail: Ol 7m

Car: 0,074 m

hindfoot: 0,230 m

shoulder height: 0,520 m

State of preservation: fair, mane almost lost

Further material of same individual: ——

Remarks: a good painting of specimen was made about 1908 by S. Ekblom

References: Renshaw (1904); Ridgeway (1909); Lénnberg (1910); Antonius

(1931)

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA (a:

STUTTGART— Skull

Catalogue number: 16884

Sex: female

Locality: ——

Date of acquisition: 1827

Remarks on acquisition: donated by Baron C. F. H. von Ludwig of Cape Town,

probably belonging to skin which was sent to Darmstadt in 1830

State of preservation: complete

References: Hilzheimer (1912); Antonius (1951); Sheer in Jit. (1959)

TRING— mounted skin

Fic. 19. Tring quagga.

Catalogue number: 394830

Sex: appears to be female; two large nipples present, vulva skin forming hump

Locality: ——

Date of acquisition: 1889

Remarks on acquisition: bought by Lord Rothschild from Edw. Gerrard,

London

History of mount: re-mounted by Edw. Gerrard before 1889

78 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description of striking features: face fairly striped; light flanking bands of dorsal

median stripe much interrupted; ventral median stripe starting at last

neck stripe

Measurements: head—body: 2°37 ia

tail: 0,37 m

ear: 0,18 m

hindfoot, left: 0,48 m

hindfoot, right: 0,47 m

shoulder height: 1,19 m

State of preservation: good

Further material of same individual; skeleton was mounted at Amsterdam in

1855—now lost (Tun 1966)

Remarks: exhibited

References: Sclater (1901); Pocock (1904); Ridgeway (1909); Griffini (1913);

Antonius (1931); Tuijn (1966)

TUBINGEN— skull

Catalogue number: 488

Sex: male

Locality: ——

Date of acquisition: 1842-50

Remarks on acquisition: Baron C. F. H. von Ludwig sent a quagga from South

Africa to Tiibingen in 1827 (Hilzheimer 1912)

State of preservation: occipital region missing

Remarks: The female quagga skull at Stuttgart is catalogued as: ‘1827 Von

Ludwig, Kapkolonie’ which leaves the origin of the Tiibingen skull

unknown, although it may possibly be one of the two quaggas formerly

in the menagerie of Friedrich I, King of Wiirttemberg

TURIN— mounted skin and skull

Catalogue number: 295?

Sex: female

Locality: ——

Date of acquisition: 1827

Remarks on acquisition: specimen bought from dealer S. Leadbeater, London;

it is possible that this is the first of the three quaggas received by the Rijks-

museum van Natuurlijke Historie at Leiden between 1827 and 1833 from

their agent at Cape Town, Dr H. B. van Horstok

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 79

History of mount: original mount with several old repairs especially on head and

lumbar region

Description of striking features: face clearly striped; light flanking bands of

dorsal median stripe, where visible, not interrupted

Fic. 20. Turin quagga.

Measurements: head—body: 2,45 m

tail: 0,47 m

ear: 0,17 m

hindfoot: 0,49 m

shoulder height: 1,19 m

State of preservation: fair, rather dirty, with rusty patches on legs

Further material of same individual: skull (No 295) in collection

Remarks: not exhibited; stored in glass case

References: Camerano (1902, 1908); Renshaw (1904); Ridgeway (1909);

Hilzheimer (1912)

VIENNA — mounted skin

Catalogue number: NMW-—St. 710 (old number 1834/II/10)

Sex: female

Locality: ——

80 ANNALS OF THE SOUTH AFRICAN MUSEUM

Date of acquisition: April 1834

Remarks on acquisition: bought from C. F. Ecklon, Hamburg; Ecklon is known

mainly for collecting botanical specimens in the Cape Colony, from where

he returned to Hamburg several times

History of mount: original mount, like Paris specimen, mainly of wood

Description of striking features: specimen is the biggest of all preserved skins

and obviously stretched as can be seen from the unusual width of both

dorsal and ventral median stripes. Face clearly striped; light flanking bands

of dorsal median stripe indicated only in anterior portion; in lumbar region

light dots unite to form irregular continuous bands; stripe fragments below

fork of shoulder stripe; body extensively striped; hair from striped areas

with light and dark transverse sections

Fic. 21. Vienna quagga.

Measurements: head—body: 2,490 m

tail: 0,420 m

ear: 0,185 m

hindfoot, left: 0,510 m

hindfoot, right: 0,530 m

shoulder height: 1,260 m

State of preservation: good, fading noticeable on left side of head and neck

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 8]

Further material of same individual: ——

Remarks: not exhibited

References: Lorenz (1902); Renshaw (1904); Lydekker (1904); Ridgeway

(1909); Griffini (1913); Antonius (1931)

WIESBADEN— mounted skin

Fic. 22. Wiesbaden quagga.

Catalogue number: 442

Sex: male

Locality: ——

Date of acquisition: 1865

Remarks on acquisition: bought from dealer Frank, Amsterdam; probably third

London Zoo quagga (Sclater 1901)

History of mount: original mount

82 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description of striking features: face faintly striped; body stripes with light dots;

dorsal portion of body striping continuing transversely almost to root of

tail, resembling gridiron pattern of Equus zebra; light flanking bands of

dorsal median stripe sometimes slightly interrupted; tail-brush thick and

very long; upper edge of hooves covered with long hair

Measurements: head—body: 2,30 m

tail: 0,41 m

ear: 0,16 m

hindfoot: 0,49 m

shoulder height: 1,20 m

State of preservation: fair; shrinkage has caused the opening of seams and

several cracks in the skin, especially of the right side.

Further material of same individual: it is considered that the skeleton at the

British Museum (Natural History), London, belongs to this individual

Remarks: exhibited

References: Anon (1858); Ridgeway (1909); Hilzheimer (1912); Antonius (1931)

Fic. 23. Doncaster; apparent quagga-hybrid foal.

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 83

Fic. 24. Dorsal patterns of: A & B. Plains zebras from Zululand; note narrow anterior portion

of dorsal median stripe and its indicated flanking bands. C. Tring quagga; note broader median

stripe and its intensified flanking bands. D. Vienna quagga; animal appears stretched through

taxidermy, which accounts for the excessive width of the dorsal median stripe; flanking bands

further intensified. E. Munich quagga; note the continuous flanking bands. F. Wiesbaden

quagga; note unusual extent of transverse stripes on rump.

= — OO

tee a

"* * QSBOIOUI OAISSAOONG : peel OL “Sz OL |

WOLVaua

‘00Z Ulllog ye seiqoz surejd uBolyy Iseq om} ‘puelnynZ ul seiqoz surejd 9014} /BIQ9Z S.[[ayoIng

g[elwuajy ZuIeY :essenb ojew zurepy ‘Teoj eddenb uMOT odea ‘Ye] Wor Sdnosd viqoz sured 94} JO SISQUWIDU Ul Sadi4ys JO oseasDep PAISSIIONG “SZ ‘Ol

ANNALS OF THE SOUTH AFRICAN MUSEUM

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 8

ACKNOWLEDGEMENTS

I should like to express my gratitude to the Director, Dr T. H. Barry, and

the Board of Trustees of the South African Museum for making my examination

of most of the preserved quagga skins possible, and thanks to those scientists

whose help made this revised list possible.

Checking old records required much patience and I should therefore like

to thank especially Dr A. W. Gentry of the British Museum (Natural History),

London; Dr A. C. van Bruggen Department of Systematic Zodlogy, University

of Leiden; Professor Dr L. B. Holthuis of the Rijksmuseum van Natuurlijke

Historie, Leiden; Dr J. Opperman of the Naturkunde Museum, Berlin; Dr F.

Fedele, Instituto e Museo di Anthropologia e Etnografia, Turin; Dr V. Eisen-

mann, Institut de Paléontologie, Paris and Dr S. Clarke of the Royal

Scottish Museum at Edinburgh. I am indebted to Dr P. A. Hulley for valuable

help in preparing the manuscript and Mrs P. Eedes for typing it.

For photographs supplied I thank the Zodlogisch Museum, Amsterdam;

the Naturhistorisches Museum, Basle; the Museum fiir Naturkunde, Berlin;

the Museum and Art Gallery, Doncaster; the Royal Scottish Museum, Edin-

burgh; the Rijksmuseum van Natuurlijke Historie, Leiden; the Naturhistorische

Museum, Mainz; the Naturhistoriska Riksmuseet, Stockholm and Mr K.

Stammerjohann of Durban.

REFERENCES

ALLEN, G. M. 1945. Quagga. Jn: HARPER, F. Extinct and vanishing mammals of the Old

World. Spec. Publs Am. Comm. int. wild Life Prot. 12: 334-339.

ALLEN, G. M. 1939. A checklist of African mammals. Bull. Mus. comp. Zool. Harv. 83: 1-763.

ANON, 1858. The quagga in the Gardens of the Zoological Society, Regent’s Park. J//. Lond.

News 33: 427-428.

ANSELL, W. F. H. 1967. Perissodactyla. Jn: MEESTER, J., ed. Smithsonian Institution preliminary

identification manual for African mammals. 6: 1-26. Washington.

ANTONIUuS, O. 1931. Zur genaueren Kenntnis des echten Quaggas (Equus quagga quagga Gm.).

Zool. Gart. Lpz. (N.F.) 4: 93-115.

ANTONIUS, O. 1951. Die Tigerpferde: die Zebras. Monogrn Wildsdugetiere 11: 1-148.

BRYDEN, H. A. 1889. Kloof and Karroo: sport, legend, and natural history in Cape Colony...

London: Longmans, Green.

CABRERA, A. 1936. Subspecific and individual variation in the Burchell zebras. J. Mammal.

17: 89-112. ;

CAMERANO, L. 1902. Materiali per lo studio delle zebre. Atti Accad. Sci. Torino 37: 612-624.

CAMERANO, L. 1908. I] quagga del Museo Zoologico di Torino. Atti Accad. Sci. Torino 43:

562-565.

Cooke, H. B. S. 1943. Cranial and dental characters of the recent South African Equidae.

S. Afr. J. Sci. 40: 254-257.

DeEsMAREST, A. G. 1820. Mammalogie, ou description des espéces de mammiferes: 414. In:

Encyclopédie méthodique. Paris.

Dorst, J. 1952. Notice sur les spécimens naturalisés de mammiféres éteints existant dans les

collections du Muséum. Bull. Mus. natn. Hist. nat., Paris (2) 24: 63-78.

ELLERMAN, J. R., MORRISON-ScOTT, T. C. S. & HAYMAN, R. W. 1953. Southern African mam-

mals, 1758-1951: a reclassification. London: British Museum.

ELorr, G. 1966. The passing of the true quagga and the little Klibbolikhonnifontein Burchell’s

zebra: which is to be next, the Cape mountain zebra or Wahlberg’s zebra of Zululand?

Tydskr. Natuurwet. 6: 193-207.

86 ANNALS OF THE SOUTH AFRICAN MUSEUM

FFOLLIOTT, P. & LiveRSIDGE, R. 1971. Ludwig Krebs, Cape naturalist to the King of Prussia;

1792-1844. Cape Town: Balkema.

FLower, S. 8S. 1929 List of vertebrated animals exhibited in the Gardens of the Zoological Society

of London, 1828-1927. 1. Mammals. London: Zoological Society.

Gray, J. E. 1850. Gleanings from the menagerie and aviary at Knowsley Hall. Hoofed quadrupeds.

[With descriptions of the specimens by Lord Derby.] Knowsley: Privately printed.

GRIFFINI, A. 1913. Le zebre. Milano: Hoepli.

Harris, W. C. 1840. Portraits of the game and wild animals of southern Africa. London:

Pickering.

HILZHEIMER, M. 1912. Die in Deutschland aufbewahrten Reste des Quaggas. Abh. senckenb.

Naturforsch. Ges. 31: 83-105.

HILZHEIMER, M. 1930. Das KOnigsberger Quagga. Z. Saugetierk. 5: 86-95.

JACKSON, A. DE J. 1920. Manna in the desert. Johannesburg: Christian Literature Depot.

LONNENBERG, E. 1910. Quagga. Fauna Flora, Upps. 1910: 85.

LORENZ, L. von. 1902. On the specimen of the quagga in the Imperial Museum of Natural

History, Vienna. Proc. zool. Soc. Lond. 1902 (1) : 33-38.

Loticuius, A. 1912. Aus der Schausammlung. Das Quagga. Ber. senckenb. naturforsch. Ges.

43: 104-107.

LUNDHOLM, B. 1951. A skull of the true quagga (Equus quagga) in the collection of the Transvaal

Museum. S. Afr. J. Sci. 47: 307-312.

LYDEKKER, R. 1904. Note on the skull and markings of the quagga. Proc. zool. Soc. Lond.

1904 (1): 426-431.

LYDEKKER, R. 1912. The horse and its relatives. London: Allen.

MEESTER, J., DAvis, D. H. S. & COETZEE, C. G. 1964. An interim classification of southern

African mammals. Zoological Society of Southern Africa & C.S.ILR.

OPPERMANN, J. 1970. Das Quagga—eine ausgerottete Zebraform. Wiss. Z. Humboldt-Univ.

Berl. (Math.-nat.) 19: 149-152.

Pocock, R. I. 1904. The Cape Colony quaggas. Ann. Mag. nat. Hist. (7) 14: 313-338.

Pocock, R. I. 1907. Notes on the quagga and Burchell’s zebra in the Paris Museum. Ann. Mag.

nat. Hist. (7) 19: 516-520.

RENSHAW, G. 1904. Natural history essays: 169-197. London; Manchester: Sherratt & Hughes.

RENSHAW, G. 1909. Zebras and quaggas. Field 11: 89.

RENSHAW, G. 1935. The quagga. J. Soc. Preserv. Fauna Emp. (n.s.) 26: 23-28.

RIDGEWAY, W. 1905. The origin and influence of the thoroughbred horse. Cambridge: University

Press.

RIDGEWAY, W. 1909. Contributions to the study of the Equidae. II. On hitherto unrecorded

specimens of Equus quagga. Proc. zool. Soc. Lond. 1909: 563-586.

RIDINGER, J. J. 1768. Das . . . nach Original-Zeichnungen geschilderte Thier-Reich. 1. Theil,

mit dem Bildn. des. sel. J. El. Ridinger. Augsburg.

Roserts, A. 1951. The mammals of South Africa. Johannesburg: Central News Agency.

Roux, J. 1910. Notes sur quelques zébres du Muséum d’Histoire Naturelle de Bale. Revue

suisse Zool. 18: 917-927.

RZASNICKI, A. 1949. Complete list of the specimens of skeletons and skins of Equus quagga

quagga (Gm). preserved in museums of the whole world in 1939. Annls Mus. zool. pol.

14: 69-73.

SCHEER, G. 1959. [Letter to E. Trumler, filed in Zoologische Sammlung des Bayerischen Staates,

Munich]

SCHOLTZ, J. du P. 1941. Uit die geskiedenis van die naamgewing van plante en diere in Afrikaans.

Kaapstad: Nasionale Pers.

SCHWARZ, E. 1912. Beitrage zur Kenntnis der Zebras. Arch. Naturgesch. 78 (A): 34-57.

SCLATER, P. L. 1901. Specimens of the quagga (Equus quagga) that have lived in the Society’s

Menagerie. Proc. zool. Soc. Lond. 1901 (1): 165-166.

SHORTRIDGE, G. C. 1934. The mammals of South West Africa. 1: 397-411. London: Heinemann.

SORDELLI, F. 1909. Note su alcuni vertebrati del Museo Civico di Milano. VIII. I] quagga

(Equus quagga Gmel.) Atti Soc. ital. Sci. nat. 48: 35-42.

TROUESSART, E.-L. 1906. La cougga et le zébre de Burchell de la collection du Muséum. Bull.

Mus. natn. Hist. nat., Paris. 1906: 449-452.

REVISED LIST OF THE EXTINCT CAPE COLONY QUAGGA 87

TusIN, P. 1966. Historical notes on the quagga (Equus quagga Gmelin, 1788; Mammalia,

Perissodactyla) comprising some remarks on Buffon-editions published in Holland.

Bijdr. Dierk. 36: 75-79.

VAN BRUGGEN, A. C. 1959. Illustrated notes on some extinct South African ungulates. S. Afr.

J. Sci. 55: 197-200.

WAGNER, J. A. 1835. Die Sdugthiere in Abbildungen nach der Natur ...von J. C. D. von

Schreber . . . Fortgesetzt von J. A. Wagner. 6: 209-215. Erlangen: Palm.

Weir, H. 1858. The quagga in the Gardens of the Zoological Society. I//. Lond. News. 33:

427. [Drawing. |

WILLOUGHBY, D. P. 1966. The vanished quagga. Nat. Hist. 75 (2): 60-63.

YORKSHIRE PHILOSOPHICAL SOCIETY. 1841. Annual report.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in paren-

theses, volume number, part number (only if independently paged) in parentheses,

pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FISCHER, P.-H., DUvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Koun, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konun, A. J. 1960b. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn: SCHULTZE, L. -

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

R. E. Rau

REVISED LIST OF THE PRESERVED

MATERIAL OF THE EXTINCT CAPE COLONY

QUAGGA, EQUUS QUAGGA QUAGGA (GMELIN)

OLUME 65 PART 3 JULY 1974

mee La

OF THE SOUTH AFRICAN

SEUM |

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 Band

July 1974 Julie

Part 3 Deel

EWO: NEW SPECIES OF ACARTIA

(COPEPODA, CALANOIDA)

FROM SOUTH AFRICAN ESTUARIES

ALLAN D. CONNELL

JOHN R. GRINDLEY

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van. stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 21, 3, 5, 7-8), 31-2, 5, t.—p.i.), 51-2, 5, 7-9),

6(1, t.-p.i.), 7(1-3), 8, 9(1-2), 10(1),

11(1-2, 5, 7, t.—p.i.), 24(2), 27, 31(1-3), 33

Price of this part/Prys van hierdie deel

R2,10

Trustees of the South African Museum © _ Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 42 9

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

TWO NEW SPECIES OF ACARTIA (COPEPODA, CALANOIDA) FROM

SOUTH AFRICAN ESTUARIES

ALLAN D. CONNELL

National Institute for Water Research, P.O. Box 395, Pretoria, South Africa

JOHN R. GRINDLEY

Port Elizabeth Museum, Humewood, Port Elizabeth, South Africa

(With 27 figures and 1 table)

[Ms. accepted 5 July 1973]

CONTENTS

PAGE

Introduction : ; : ; A . : 89

DEescripuonolsmatenialy we) as glee oe 89

Summary . ; : : : : : : 97

Acknowledgements : , ; 3 f 97

References , : . ; : ; Fs 97

INTRODUCTION

During studies of the zooplankton of South African estuaries two species

of Acartia new to science have been found. One species, referable to the sub-

genus Acartiella, is often the numerically dominant copepod in estuaries on the

east coast of southern Africa. The other species, of the subgenus Paracartia,

occurs in estuaries on the east coast of South Africa, usually in smaller numbers

although it is abundant occasionally. Acartiella and Paracartia are here regarded

as subgenera (cf. Wellershaus 1969) despite Gurney’s (1931: 217) proposal to

remove them from Acartia. Bowman (1965: 149) pointed out that Steuer’s

(1915, 1923) primary division into ‘Acartiae arostratae’ and ‘Acartiae rostratae’

was unacceptable although his subgenera may be maintained.

DESCRIPTION OF MATERIAL

Acartia (Acartiella) natalensis sp. nov.

Figs 1-12, 24, 25

Material examined

Numerous specimens in samples from estuaries between Knysna (34°S)

and the Morrumbene estuary, Mozambique (23°S).

Types

Holotype male (Reg. No. SAM A13408), allotype female (Reg. No.

SAM A13409) and paratypes 1034, 1099 (Reg. No. SAM A13410) deposited

in the South African Museum, Cape Town, from the Mtentu River estuary

(31°14,5’S, 30°2’E) on the Pondoland coast of South Africa.

Ann. S. Afr. Mus. 65 (3), 1974: 89-97, 27 figs, 1 table.

90 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description

Female (Figs 1, 3, 6-9, 11)

Total length 0,93 to 1,05 mm (Morrumbene specimens were considerably

smaller: 0,87 to 0,90 mm, and Knysna specimens were also smaller: 0,82 to

12b

12¢

Figs 1-12. Acartia (Acartiella) natalensis sp. nov. 1, female dorsal view; 2, male, dorsal view;

3, female, urosome, dorsal view; 4, same, male; 5, geniculate portion of male right first antenna;

6, female, second antenna; 7, female, mandible; 8, female, maxilliped; 9, first leg, female;

10, male, fourth leg; 11, female, fifth leg; 12, male, fifth leg (posterior view); 12a, 12c, anterior

view, right leg; 12b, anterior view, left leg. All measurements in microns.

TWO NEW SPECIES OF ACARTIA 9]

0,90 mm). The head is slightly produced anteriorly. No rostrum or filaments

are present. The postero-lateral corners of the prosome are spineless (Fig. 1).

The first antennae reach the posterior end of the furcal rami. Urosome segments

are devoid of spines (Fig. 3), but the anal segment usually bears a single trans-

verse row of minute hairs on its dorsal surface. The furcal rami each bear five

long plumose setae, four terminal and one dorsal, and a short curved accessory

seta laterally about twice as long as ramal width. The furcal rami are asymmetri-

cal, the right being longer.

The first antennae consist of 22 apparent segments, but ‘segment’ 2 is

apparently compounded of 3 segments and ‘segment’ 4 of 2, giving 25 true

segments. Segments 13 to 20 (apparently 10 to 17) have a row of small spines,

slightly less in length than the width of the segments. Segment 22 is only frac-

tionally longer than wide.

The second antennae are somewhat atypical for the subgenus, being of

the normal Acartia form (Fig. 6), and apparently resembling those of A. (Acar-

tiella) kempi (Sewell, 1914), the only other described member of the subgenus

with this type of second antenna. The endopod is relatively small. There are no

hairs on the inner margin of the exopod. There are slight differences in setation

in specimens from different localities.

The first maxilla has the base more expanded than in that of A. (Paracartia)

longipatella sp. nov. (Fig. 21) and lacking the patches of fine setae on the inner

margin. There are no hairs on the distal margin of the exopod.

The second maxillae are normal, bearing approximately 15 setae of varying

length. The maxillipeds are as figured (Fig. 8) with three strong and two reduced

biplumose setae on the basal segment. Specimens from Knysna show these five

setae more equally developed. The four pairs of swimming legs are typical of

the genus (Table 1). On the first leg the second segment of the exopod, however,

lacks the outer marginal spine on its apical angle (Fig. 9), while this spine is

present but much reduced on legs 2 to 4. This spine is barely noticeable in legs

2 and 3 in many specimens. Leg 5 is distinctive from that of all other known

species of the subgenus Acartiella. The exopod is basally about four times

thicker than the endopod (Fig. 11), while the endopod is ? as long as the exopod,

bearing short hairs as figured. There is no seta midway on the outer margin of

the exopod.

Male (Figs 2, 4, 5, 10, 12, 24, 25)

Total length 0,82—0,88 mm (Morrumbene specimens 0,76-0,79 mm, Knysna

specimens 0,73-0,78 mm). General shape is similar to the female although the

urosome is longer relative to the prosome length (Fig. 2). The last urosome

segment has a couple of small spines, and there are usually two or three on the

outer edge of the caudal rami (Fig. 4). Only the left ramus has a short curved

accessory seta, while the plumose setae are similar to those of the female.

The first antennae extend to the distal end of the furcal rami. The left

antenna is as in the female, while the right has only 17 apparent ‘segments’,

92 ANNALS OF THE SOUTH AFRICAN MUSEUM

Figs 13-23. Acartia (Paracartia) longipatella sp. nov. 13, female, dorsal view; 14, male, dorsal

view; 15, female, urosome, semi-schematic, showing position of accompanying plates; 15a,

same, lateral view; 16, dorsal view male urosome; 17, female, rostrum, and filaments; 18,

male, right first antenna; 19, male, second antenna; 20, male, mandible and palp; 21, female,

first maxilla; 22, female, fifth leg; 23, male fifth leg, anterior view; 23a, another detailed view

of the left fifth leg. All measurements in microns.

TWO NEW SPECIES OF ACARTIA 93

TABLE 1

Ornamentation of the swimming legs of Acartia (Acartiella) natalensis sp. nov.

Si, Se, St represent internal, external and terminal spines or setae respectively.

The number of setae is shown in arabic numerals and spines in roman numerals.

Beto 0:| 0 0

* much reduced

** reduced

several of them compound. The geniculation is between ‘segments’ 15 and 16

these being compounds of true segments 17-18 and 19-21 respectively (Fig. 5).

Apparent segment 9 bears a short heavy spine, while 10 and 11 are swollen.

‘Segments’ 14, 15 and 16 bear serrated, sabre-like spines (Fig. 5).

The remaining appendages from the second antenna to leg 4 are as in the

female.

The male 5th legs differ markedly from those of other members of the

subgenus (Figs 12, 12a, 12b, 12c, 24, 25). The right leg does however form a

well-developed clasping apparatus as specified by Sewell (1914) in his definition

of the subgenus. The detail of the terminal structures of P5 left segment 2—3 was

difficult to determine clearly and was therefore photographed under the scanning

electron microscope after critical-point drying (Figs 24, 25).

Occurrence

This species was abundant in most estuaries on the east coast of southern

Africa between Knysna and Morrumbene. It was collected in waters of varying

salinity from 3,3 to 65%.

Discussion

Although the shape of the second 2 %tenna places this species with kempi

Sewell, 1914, the male and female 5th leg; are unusual and somewhat different

from any known species. The key to Acartiella species provided by Wellershaus

(1969) (from which A. sinensis Chia-jui & Foo-siang, 1963 is omitted) easily

accommodates natalensis, since this species is easily distinguished from kempi

by the structure of the 5th leg of both male and female.

94 ANNALS OF THE SOUTH AFRICAN MUSEUM

Figs 24-25. A. (Acartiella) natalensis sp. nov. Male fifth leg, left, 1, lateral view, 2, dorsal view

(with some debris caught in the setae).

Figs 26-27 A. (Paracartia) longipatella sp. nov. Male fifth leg, left, lateral view.

TWO NEW SPECIES OF ACARTIA 95

This species is one of the dominant copepods at least as far north as the

Morrumbene estuary, suggesting a wider distribution probably extending to

estuarine waters of the east African coast. Its presence in South African waters

considerably increases the known range of the subgenus Acartiella, previously

confined to Indian, Burmese and south Chinese coastal and estuarine waters.

Acartia (Paracartia) longipatella sp. nov.

Figs 13-23, 26, 27

Material examined

Many specimens in samples from estuaries from between the Klein River

estuary (Hermanus) and St. Lucia (Natal). It was not observed in samples from

Mozambique estuaries or from estuaries on the west coast of South Africa.

Types

Holotype male (Reg. No. SAM A13411), allotype female (Reg. No. SAM

A13412) and paratypes 1039, 1099 (Reg. No. SAM A13413) deposited in the

South African Museum, Cape Town, from the Mtentu River estuary (31°14.5’S,

30°2’E) on the Pondoland coast of South Africa.

Description

Female (Figs 13, 15, 17, 21, 22)

Total length 0,98—1,04 mm (Knysna specimens 0,80—-0,88 mm, Breede River

specimens 0,82-0,90 mm). The head is slightly produced anteriorly. A rostrum

and filaments are present (Fig. 17). The postero-lateral corners of the prosome

are rounded (Fig. 13), with the first urosome segment greatly enlarged, the

posterior two segments normal (Fig. 15). The caudal rami are short, only

slightly longer than wide. Caudal setae are as figured (Fig. 15). On mated females

the accompanying plates (Sars 1904) of the spermatophore reach far past the

posterior tip of the furcal rami (Fig. 15).

The first antennae reach almost to the posterior end of the first urosome

segment (Fig. 13), with 19 recognizable segments. The second antennae are of

the normal Acartia form, with fine setae on the inner margin of the exopod and

second basal segment (Fig. 19). The mandible with its palp is as figured for the -

male (Fig. 20). The first maxilla is figured (Fig. 21). The second maxilla is typical

of the genus, bearing approximately fifteen setae of varying length. The

maxilliped is similar to that figured for A. (Acartiella) natalensis (Fig. 8), but

the second segment is more swollen and has its plumose spine much shorter

than that of A. natalensis.

The first swimming leg is similar to that of A. natalensis (Fig. 9) but has

an outer marginal spine on the apical angle of each of the three exopod segments,

while segment 3 has, in addition, a median external seta and a short spine near

the base of the terminal spine. The setal formula of the inner margin of the

96 ANNALS OF THE SOUTH AFRICAN MUSEUM

endopod segments is 1:3 compared with 0:2 of A. natalensis (Fig. 9 and Table 1).

Legs 2-4 are similar to A. natalensis (Fig. 10), but a short stout spine is well

developed on all 3 segments of the exopod.

The female fifth legs resemble those of other Paracartia species, but the

outer plumose setae are subequal in length to the exopods (Fig. 22), which are

not as strongly curved as those of P. africana (Steuer).

Male (Figs 14, 16, 18-20, 23, 26, 27)

Total length 0,82-0,88 mm (Knysna specimens 0,76-0,87 mm, Breede

River specimens 0,75-0,79 mm). The head is rather more truncate anteriorly.

The last segment of the prosome has rounded postero-lateral corners (Fig. 14).

Urosome segments are not greatly enlarged. The furcal rami are no longer than

their width. Setae are as illustrated (Fig. 16).

The first antennae reach to the posterior margin of the prosome (Fig. 14).

The left antenna is as described for the female, the right is as figured (Fig. 18),

and the geniculation is between the fourteenth and fifteenth ‘segments’, the

fourteenth bearing a stout, blunt spine slightly longer than segment length.

‘Segments’ 9-13 are hollowed anterodorsally.

The appendages from second antenna to fourth leg are as for the female.

The right fifth leg is greatly enlarged (Fig. 23), and is typical of males of the

subgenus Paracartia (see Steuer 1923). The left side is also fairly typical, being

somewhat reduced. The terminalia include a long thin spine and a broader

plate, setose along one edge (Figs 23, 23a). The line-drawings were confirmed

by scanning electron microscope photographs (Figs 26, 27).

Occurrence

This species was present in estuaries between Hermanus (Klein River

estuary) and Natal (St. Lucia), usually in small numbers but occasionally abun-

dant. It was obtained in waters of varying salinity from 7—35,5%,.

Discussion

There are only four members of the subgenus previously described and their

distribution ranges down the east side of the Atlantic Ocean, with P. /atisetosa

(Kriczagin) penetrating into the Mediterranean and Black Seas. The most

northerly species is P. grani Sars, from the north-west coast of Europe (Sars

1904), while P. dubia Th. Scott has been collected in the Gulf of Guinea and

P. africana along the South West African coast (Unteriiberbacher 1964) and

west coast of South Africa (unpublished record).

Acartia asymmetrica Tanaka, 1964, described from the ‘Bay of Cape Town’

is a synonym of Acartia (Paracartia) africana Steuer.

The occurrence of a member of the subgenus in east coast temperate and

subtropical estuaries is thus interesting, and extends the range of the subgenus

into the Indian Ocean.

The lack of any form of wing-like projection on the last prosome segment

TWO NEW SPECIES OF ACARTIA OF

in the female is distinctive, as is the shape of the first urosome segment and the

size and shape of the spermatophore plates, in mature females. The prominence

of these long plates gives rise to the name proposed for this species.

Both Acartia (Acartiella) natalensis and Acartia (Paracartia) longipatella

appear to be restricted to estuaries and neither species has been found in the

open sea.

SUMMARY

Two new species of Acartia are described from estuaries on the east coast

of southern Africa. Acartia (Acartiella) natalensis occurs in estuaries in South

Africa and Mozambique, while Acartia (Paracaria) longipatella has only been

recorded from estuaries on the east coast of South Africa.

ACKNOWLEDGEMENTS

We thank Mr T. R. Wooldridge and Mr A. F. de Villiers for assistance with

the examination of specimens. We gratefully acknowledge that Dr A. de Decker

of the Division of Sea Fisheries also recognized these two species as undescribed

but left their description to us.

REFERENCES

Bowman, T. E. 1965. An arostrate population of the copepod Acartia lilljeborgii Giesbrecht

(Calanoida, Acartiidae), from St. Lucia, West Indies. Crustaceana 8: 149-152.

CuiA-Jul, S. & Foo-SIAna, L. 1963. The estuarine Copepoda of Chiekong and Zaikong rivers,

Kwangtung Province, China. Acta zool. sin. 15: 571-596.

GURNEY, R. 1931. British fresh-water Copepoda. 1. Ray Soc. Publs 118: 1-238.

Sars, G. O. 1904. Description of Paracartia grani G. O. Sars, a peculiar calanoid occurring

in some of the oysterbeds of western Norway. Bergens Mus. Arb. 1904 (4): 1-16.

Scott, T. 1894. Report on the Entomostraca from the Gulf of Guinea. Trans. Linn. Soc.

Lond. (2) 6: 1-161.

SEWELL, R. B. S. 1912. Notes on the surface-living Copepoda of the Bay of Bengal I and II.

Rec. Indian Mus. 7: 313-382.

SEWELL, R. B. S. 1914. Notes on the surface Copepoda of the Gulf of Mannar. Spolia zeylan.

9: 191-263.

SEWELL, R. B. S. 1919. A preliminary note on some new species of Copepoda. Rec. Indian Mus.

16: 1-18.

STEUER, A. 1915. Revision der Gattung Acartia Dana. Zool. Anz. 45: 392-397.

STEUER, A. 1923. Bausteine zu einer Monographie der Copepodengattung Acartia. Arb. zool.

Inst. Univ. Innsbruck 1: 89-147.

STEUER, A. 1934. Two new copepods of the genus Acartia from Burma. Rec. Indian Mus. 36:

335-338.

TANAKA, O. 1964. Two small collections of copepods from the Antarctic. Scient. Rep. Jap.

Antarct. Res. Exped. (E) 22: 1-20.

UNTERUBERBACHER, H. K. 1964. Zooplankton studies in waters off Walvis Bay with special

reference to the Copepoda. Invest] Rep. mar. Res. Lab. S.W.Afr. 11: 3-42.

WELLERSHAUS, S. 1969. On the taxonomy of planktonic Copepoda in Cochin backwater (a

South Indian estuary). Veréff. Inst. Meeresforsch. Bremerh. 11: 245-286.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BuLLouau, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FIscHer, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FISCHER, P.-H., DuvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konan, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konan, A. J. 19605. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

Allan D. Connell & John R. Grindley

TWO NEW SPECIES OF ACARTIA

(COPEPODA, CALANOIDA)

FROM SOUTH AFRICAN ESTUARIES

_ VOLUME 65 PART 4 JULY 1974

ae 7. A

‘OF THE SOUTH AFRICAN.

ROVN 2 KO) WANE

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 Band

July 1974 Julie

Part 4 Deel

ON THE TAXONOMIC STATUS, DISTRIBUTION

aD ECOLOGY OF THE BLUE ANTELOPE,

mere OLRAGUS LEUCOPHAEUS (PALLAS, 1766)

RICHARD G. KLEIN

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 2, 3, 5, 7-8), 3(1-2, 5, t.-p.i.), 51-2, 5, 7-9),

6(1, t.—p.i.), 71-3), 8, 911-2), 10(1),

11(1-2, 5, 7, t.—p.i.), 24(2), 27, 31(1-3), 33

Price of this part/Prys van hierdie deel

R6,00

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 43 7

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

ON THE TAXONOMIC STATUS, DISTRIBUTION AND ECOLOGY OF

THE BLUE ANTELOPE, HIPPOTRAGUS LEUCOPHAEUS

(PALLAS, 1766)

RICHARD G. KLEIN

Department of Anthropology, University of Chicago, Chicago

(With 2 figures, 1 map and 3 tables)

[MS. accepted 10 July 1973]

CONTENTS

PAGE

Introduction : ; : 5 A i ' ; : 99

Definition of the samples ; : : ; ’ . 100

Definition of the measurements. : : : s 106

Analysis of the measurements i : : : . 108

Conclusions ; ; : 3 ‘ 4 : : Mitek lia

Summary . : 2 ; ' ; ; : 5 oo) Ob:

Acknowledgements . ; : ; : ; : ad.

References . é 3 ; é : . : ; OE STS

Appendices : : j 5 ‘ ; : ; ae GLhG

INTRODUCTION

The first published account of the blue antelope was made by the German

Peter Kolb (1719 as cited in Mohr 1967: 6-10) who lived and travelled in what is

now known as the south-western and southern Cape between 1705 and 1712.

Subsequently, other nineteenth-century visitors to the Cape published indepen-

dent descriptions of this creature, which they encountered just east of the

Hottentots-Holland Mountains, mainly in the triangle Swellendam—Caledon-—

Bredasdorp but occasionally as far east as Plettenberg Bay. On the basis of skins

and skulls sent back to Europe, Pallas (1766 as cited in Mohr 1967: 11) presented

the first truly systematic description of the species, which he called Antilope

leucophaea. In 1774 Thunberg reported that the blue antelope had become very

uncommon (Mohr 1967: 6). The last one was seen around 1800, making this

species the first historically recorded African mammal to become extinct

(Harper 1945: 698-700).

The early extinction of the blue antelope, before qualified scientists could

observe wild or even captive specimens, left open many questions concerning its

Ann. S. Afr. Mus. 65 (4), 1974: 99-143, 2 figs, 1 map, 3 tables.

100 ANNALS OF THE SOUTH AFRICAN MUSEUM

physical appearance, taxonomic status, geographic distribution, and ecology.

There are, of course, no photographs, and most of the available sketches and

descriptions are obviously inaccurate in one respect or another. Kolb, for

example, drew the creature with a beard, apparently because he thought it was a

close relative of the goats (he placed the blue antelope in the genus Capra) and

some subsequent writers followed this custom. From Mohr’s (1967) review of

the early drawings and descriptions, it is apparent that they often contradict one

another and are also at variance with the four mounted specimens still available

in Europe, as well as with what may be surmised about the blue antelope from

knowledge of its closest living relatives, the roan (Hippotragus equinus) and the

sable (H. niger).

In her monographic study of the blue antelope, Mohr (1967) was forced to

rely heavily on the mounted specimens (one each in Vienna, Stockholm, Paris,

and Leiden) and on a skull housed in the Hunterian Museum at the University

of Glasgow. This skull has no history attached to it, but Broom (1949) assigned

it to H. leucophaeus, mainly because there was reason to believe it had been

obtained before 1800—a time when the blue antelope was the only species of

Hippotragus that had been encountered by Europeans (according to Broom,

H. equinus was first seen by Europeans in 1804, H. niger in 1836). In the few

years since Mohr’s monograph was completed, a considerable amount of rele-

vant osteological material has been found at archaeological and palaeontological

localities in or near the area where H. leucophaeus was recorded historically. The

purpose of this paper is to summarize the taxonomic, distributional, and

ecological implications of this material.

DEFINITION OF THE SAMPLES

The initial stimulus for this paper was the observation that teeth assignable

to the genus Hippotragus from the archaeological site of Nelson Bay Cave

(Plettenberg Bay) were remarkably variable in size (Fig. 1), being sometimes

significantly smaller and sometimes appreciably larger than homologous teeth

of the largest H. equinus specimen in the South African Museum’s comparative

collection. Nelson Bay Cave lies within the historic area of distribution of H.

leucophaeus, and it seemed highly probable that it would be represented at the

site. With this in mind, two explanations of the observed size variation in Nelson

Bay Hippotragus teeth seemed possible: (1) All the Nelson Bay teeth derived

from H. leucophaeus, which was highly variable in size and which, contrary to

historic observations, included some individuals as large or larger than the

living roan; (2) the Nelson Bay sample included teeth from H. leucophaeus and

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 101

also from another species of Hippotragus, perhaps H. equinus In order to

determine which hypothesis was more likely and possibly at the same time

establish some of the metrical characteristics of H. leucophaeus vs. those of

H. equinus and H. niger, it was obviously necessary to obtain measurements on

samples of well-defined H. equinus and H. niger and also on fossil Hippotragus

from other southern Cape sites.

A list of the relevant fossil sites with some background data on each is

presented in Table 1. Map 1 gives their locations. At most of the sites, analysable

Hippotragus remains consist overwhelmingly of isolated teeth. Whole dentitions

are common only from Swartklip and even there the number is not large. Three

sites have provided analysable horn cores (an example from Nelson Bay Cave is

illustrated in Fig. 2). No complete or even nearly complete skulls are known.

Post-cranial remains occur at several sites, but have been ignored here because:

(1) they are exceedingly difficult to distinguish from the post-cranial bones of

other similar-sized bovid genera represented in the same collections; (2) most of

the post-cranial material is highly fragmentary, greatly reducing its descriptive

and analytic value (and compounding the difficulty of generic identification);

and (3) there are no large museum samples of well-identified recent Hippotragus

post-cranial remains with which to compare the fossil material. Data on the

quantities of analysable teeth and horn cores available from the different sites

are presented in Appendix I.

The recent samples of H. equinus and H. niger with which the fossil material

is compared are composites of collections housed in the South African Museum

(Cape Town), the Transvaal Museum (Pretoria), the National Museum of

Rhodesia (Bulawayo), the British Museum (Natural History) (London) and the

Field Museum of Natural History (Chicago). The geographic provenances of the

samples are given in Table 2, from which it can be seen that specimens from

south-central Africa (especially Zambia) predominate heavily in both samples,

giving them a distinct geographic bias. There are two few well-provenanced

specimens from other areas in either the H. equinus or the H. niger sample to say

with any certainty that geographic differences in size do not characterize either

or both species, but the data are sufficient to argue that any differences which do

exist are probably small. Further, my search of the literature has failed to turn up

any references to marked size differences among recognized subspecies of either

H. equinus or H. niger, with the exception of H. niger variani, the ‘giant sable’ of

Angola. I encountered only one H. niger variani specimen in the museum col-

lections I examined (in the Field Museum) and measurements on it were recorded

separately from those of the remaining H. niger sample. I think it is fair to

conclude that the admitted geographic bias of the two comparative samples does

not disqualify them for use in this study.

1 The possibility that the second species might be the extinct giant hippotragine, Hippotragus

gigas, was ruled out since the relevant Nelson Bay teeth were all morphologically quite distinct

from teeth of H. gigas, as known, for example, from Elandsfontein. The dentition of H. gigas,

in fact, is morphologically more like that of recent Oryx spp. (though much larger) than that

of recent Hippotragus spp.

ANNALS OF THE SOUTH AFRICAN MUSEUM

102

(UMO}SUUeYeIO)

uinosnyy Aueq ry

yosoqual[a}g Jo A}ISIOATU/)

‘KZo[OoRYyoIY JO JuoujIedoeqd

wuunesnypy 4j9qQeZI[q Od

umnasnyy UROTpY YINOS

wuunasnyy URdLIpyY YNOS

wmoesnyy UeoLIpy YNog

wmnesnyy UROLIyy YyINOS

umnesnyy UedIpy YNoS

wnesnpy UedIyy yyNOS

winesnyy URdTIyy yINOS

umnesnpy ULLIyy YNOS

umMosnyy UROIy YINOS

UMO], odeD jo A}ISIOAIUS)

‘AZo [OovyoIY JO yUouNIedaqd

Jolsajoul fo uoljvIO] Juasadd

(‘wiwI09 ‘siod) 1ayOo1g “"W

(ZL6I “6961) UooRed

(pL61) UOT *(OL6T) SIP

(pL61) Ulopy ‘(Ssoid ul pue

ZLO61) JOBUIS 2 IOWAAA

(ssoid Ur) JozIng

‘(8961) AopuoH 29 AopusH

(pL6l “9 ‘PTL6I) WOT

(ZL61) USIS[OH 2 19zZINg

(0961) STI9AA

‘(E61) 72 42 UIMpooH

QUON

(pL61) UI

(PL61)

KapueH :(L/61) 24900

(pL61) WIM

“(PL61 “6961) SopucH

(1qndun)

UlgT *(ZL6T) UOsULyIed

aouasafay

(0U9I19jUI

[eo1do[oseYyoie) USDO[OH]

(vI-D)

“d'A 009 L P410J0q APIOYs 0} OOP ST

(QoudIOJUT [ROIZO[OSeYIIe)

[eloe[D yey Josey

(gousIajUI [edIsO[ONS pure pI-D)

[eloe[s19}U] ySeT OY} JO pus

oy} pue ‘d’d 000 0€ < YseMjog

(s0uUsIOJUI

[eo1sofoe3 pue p[-D) [eloe[D

se] Joryiee {p'd'q Ssieak 000 OF<

o(PI-D) ‘d’& 000 S—008 81

(Q0USISJUI

[edISO[OSB) [eIOR[H 3seT 19}eTZ

(20UeISJUI

[eo1so[OoRYySIe) DUZDO[OH{

pure [elderly sey [eUluoL,

(eyep souPUDAOId OU) DUSD0[O Hy

(QoudIOJUI [eOIsO[OORYSIe

pue [eo1so[ojuoseyed)

QUdDO[OH{ Pur [eIDR[H 3seT JoyeT

(Qous1OJUI [edISO[O}UOSE]ed)

guss0}sio[q Ioddné

q(Q0UsrazUL [edISOTO}UOSETe)

9us00}sio[q Jodd~) pue s[pplW

(pI-D)

‘d' 009 6 Joye Apsoys 0} 000 IT

e(Q2UaplAa

Suysoddns fo aanjou pup) spissof

sn3enoddiy fo asp jv2180]0a5)

Sox

Sok

Ajqeqoid

Sox

sok

sox

A[qeqoid

AIQISsod

yied Ul

ysea] ye A[Qeqoig

Sox

é 44141490

uDUINY 0} pajDjad

uoljDjnuinJID auog

Joyous e[epuoiuy

woognoyyeW

s3ulidg yoopelD

SOARD YINOP

JOATY SOISETY

dipyemg

dAegQ Aeg UOSTON

(l9[AUSOID)

juURseo[d Dye]

ysInyyxeO

UO}SMEH

DARD S,dIAq

soquiso[g

ul} UOJspuely

oaAeD Aeg spur[q

ails

‘PoALIop oJom Joded sty} ul posAyTeue suTeUOI snsv4joddipyT 3Y} YOY WO says [eoIsoTOoeysIe pue eosojoyuoerjed adeD ussyynog

T 919%.L

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 103

*(AJOAT}O0dso1

snuinba *H] 10 snavydoonay Snspdjoddif] 0} pousisse S9UOq SUIJUOS AjITeDO] — aFY “[H) 3x9} 9} Ul psuONUSU satjITeso0] odeD UIOYNOS *| ‘dew

ole 092 SZ ob2 ofc oce old 002 06! of!

| | Tey BADD

$ O44: q

NYY JOAI, °

(IH) tLe fs Mey duopsppeig oie

obf // (ay B/H) AAD

| 40,

WOpus||aMS

--~— —-— -9----~—

UWope1056 ne

=(IH) I/yf40MS.

ss ¢ Ig U0s/g

ie = ita Bsoquosseig® 71

—_-

~ —— —

(°H 8 /H) N

4AL/OYS 5 (OH re

gate \wooqinoyyay

off TS

| °

(1H) z Oqusao/g

(IH) Urequoyspuof. Rs

$18 pepjoy edog

$0 $lWI] @yowIxouddo

—_— -_ =

TS ee

> —

(9H) \

SB6UISAS YIOPDII> x |

eze (IH) aA0D Aogspuo/z7

092 ° obd ofe ocd old 002 a 08!

“dippzemg ye souoqg

JECUUIU YIM UOleIDOSse 4O9IIP Ul [[aYS3o YOI}SO UO Psule}qo UsEq sey (“WUD ‘sad “AO[{IN) (OP89-]) “s1eO4 YOO OV< JO 9}ep PI-D JISUIS V p

“posdyeue usaq 304 OU sey Woy} WOT euney

UeTeuUeu 94} nq ‘sureWoI sUIseOddIY urle}UOS os[e ARUI OSOUL, “dG OOO § UY} IoSUNOA ore yey} Avg UOSTAN 32 syIsodap poyeavoxo sey dsoysuy »

‘IosuNOA A[QviopIsuod oq AeuI ‘snapydoong] "FT 0} pousisse souoqg

sdeyiod Surpnjour ‘uoniod v INQ “(PL6I Ule[y) ouD00}sIE1g SIPPHN SY} wosy sayep A[qeqoid euney oy) Jo ynq oY ‘Spues SUIATIDAO JO UOTyeyop

jU9901 SULINP posodxoa saovjinsosvyed SNOUISNAIAJ IO snosred[ed SIOU 10 OU WOAF P9}9][O9 Usaq svY [eIIo}eU Jeune] UlOJUOJspuL[_ ou} Jo SOW gq

‘d'@ 000 02-000 SZ

[BAIJUL OY} UI s[ey ATQeqoid Arepunoq [e1e[ Ise T/[eloe[s19}U] 4Se'] OL “dg OOO OL } “dA 000 ZI Wory [ears}UT ow oY} ATYBnoL ‘suIE} oynjosqe

UI IO ‘Teor (ULNA) ISP] Oy} pue [eloe[s10juy (Woq) yseJT oY} Jo Ueds-oUll} POUIQuIOS oY} 0} JayOr 07 e19Y PEsn SI 9U9004SIo[q Joddy) W190} oy], «

> SALON,

104 ANNALS OF THE SOUTH AFRICAN MUSEUM

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 105

(HO9§ “yf Aq ssuIMeIq) “(ZLLTO WYS) jUesee

(AW)"dP “(MAS)*dP bh “Wh S(SaE/SO WYS) eae Keg uosjan WOJJ Snapydoona *H Jo (MAW :A¢

Wory snuinba “FT JO (AW)W : az ‘Wz ‘(lab/fsa WYS) 2485 A

SS oe

ba UOSTON Woy snuinba snsosjoddifT JO (MA

Tatts

Id Oe] Wor snapydoona] “7 Jo (pajdnioun)'w “(Ma)'W “(MW)'aP

"VE ‘(LOP6/fSA WYS) Ae eg uosjon

JW “(AAW)*dP “(AAW)*aP HT CWT CT ‘SI

106 ANNALS OF THE SOUTH AFRICAN MUSEUM

Table 2

Geographic origin of the comparative specimens of Hippotragus used in this study.

NUMBER OF SPECIMENS

H. equinus H. niger subspp. H. niger variani

South Africa eat Ga

Swaziland 1

South West Africa 1

Botswana 1

Rhodesia : ; i ; : 5

Zambia . : : : : : : 45

f ’ 3 : F 2

moOONnNodcdh

Malawi

Angola ;

No provenance given in museum ‘cata-

logues . 3 : : i ‘ 39

Lillo lla} (=) (i=)

co}; =

N | \O

rlo

TOTALS: =o eer) 99

DEFINITION OF THE MEASUREMENTS

The only dimension measured on both fossil and recent teeth was maximum

length at the level of the occlusal surface, on the buccal side for maxillary teeth

and on the lingual side for mandibular ones. Whenever possible (mainly on the

comparative specimens), maximum length of whole premolar and molar rows

were also recorded. The maximum length measurement has the dual advantage

of analytic utility and easy definition and replication. Furthermore, it is the

dental measurement most frequently used by other investigators of bovid fossils

(see, for example, Gentry 1966 or 1970).

Casual observation is sufficient to show that the individual teeth, and espe-

cially the molars of Hippotragus spp., change length as they wear. In order to

reduce sample variance and possible sample bias from this factor, four wear

categories were defined and only measurements on homologous teeth in the

same wear category are compared below, except in the case of whole molar and

premolar rows where observation suggests there is relatively little length change

with wear in any case (as some teeth grow shorter within a row, others become

correspondingly longer). For the molars, the four wear categories are: (1) No

Wear (NW)—no obvious wear on the molar crown; (2) Early Wear (EW)—

wear on the crown obvious, but basal pillar not part of occlusal surface; (3)

Medium Wear (MW)—basal pillar part of occlusal surface but tooth still

characterized by considerable height above the alveolus (or on isolated teeth

above the roots); and (4) Late Wear (LW)—tooth worn down to near the level

of the alveolus (or, in the case of isolated teeth, to near the level of the roots, with

occlusal surface very flat and smooth). The same four categories were used for

the premolars (both deciduous and permanent), but are harder to define because,

with the exception of dP,, the premolars do not generally possess basal pillars.

Separation of premolars in Early Wear and in Medium Wear is thus particularly

difficult, but tests made to see whether the same tooth would be classified in the

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 107

nt Wa \

Shenae

W\ Ne

“\ KN

ae WS

\: e ‘

(oa)

Ve)

a>}

aa)

je)

ie)

(5)

io)

is)

108 ANNALS OF THE SOUTH AFRICAN MUSEUM

same wear category in successive weeks included no failures, suggesting that,

however difficult to define, wear category judgements were at least consistent.

In any case, the premolars do not change shape and length from Early Wear to

Medium Wear as dramatically as the molars do so that confusion of premolar

wear categories is unlikely to seriously affect any conclusions reached below.

Since most of the fossil horn cores lack the tip and sometimes a good part

of the length adjacent to the tip, and since the sheaths could not be completely

removed from many of the comparative specimens, horn-core measurements

were restricted to the maximum longitudinal (antero-posterior) and transverse

(medio-lateral) diameters immediately above the pedicel. These diameters are

useful not only in themselves, but also because the ratio between them is a

measure of transverse (side-to-side) compression. Thus, taken together, they

may be used to say something about both the size and the shape of a basal horn-

core cross-section. One difficulty in using the diameters to compare hippotragine

species, however, is the fact that there are very great differences in basal horn-

core size and shape between subadults and adults and between males and females

within both H. equinus and H. niger. (In both species, young females have the

smallest, most rounded horn-core bases, adult males the largest, most trans-

versely compressed ones.) Sex is easy enough to take into account, but age is

somewhat more difficult (for the teeth it is implicitly taken into account by the

use of the wear categories). Since the bony consistency (solidity and non-

porosity) of all the fossil horn cores suggested they belonged to adults or near-

adults, I decided to concentrate on measurements of obviously adult specimens

in the comparative collections. Because I could not always determine adulthood

by examination of the cores themselves (frequently I could not remove the

sheaths sufficiently), I relied on a dental state (upper and lower M3’s both in at

least Early Wear) as a rough index.

Appendix 1 presents the basic statistical parameters (xX = arithmetic mean,

N = sample size, s = standard deviation) for each dental and horn-core category

within H. equinus, H. niger, and each of the fossil samples. For H. equinus and

H. niger the data are further broken down by sex, including a category in which

known males, known females, and specimens of unknown sex are lumped. For

H. equinus, H. niger, and those (few) fossil specimens for which matching left and

right halves were available, only measurements on the left half were used to

calculate the parameters in Appendix 1. All measurements were made in milli-

metres with the same dial-reading Helios calipers.

ANALYSIS OF THE MEASUREMENTS

Although a variety of statistical procedures, including multi-variate ones,

could be used to analyse the numbers in Appendix |, it was felt that the easily

understood ‘t test’ for statistically significant differences between two means,

together with an intuitive appreciation for the size and patterning of differences,

would be adequate to deal with the questions that prompted the analysis. Values

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 109

of ‘t? between paired means were calculated using a modified version of Pro-

gramme BMDX70 on the CDC 6400 at the University of Washington computer

centre (the same programme also calculated the means, standard deviations, and

coefficients of variation found in Appendix 1). Those samples whose means were

found to differ from one another at the 0,05 significance level or below are listed

in Appendix 2. The sample with the larger mean is always to the left. In each case,

the value of ‘t’, the number of degrees of freedom (df), and the actual significance

level of the differences (p) are given in succeeding columns.

Appendix 2, in conjunction with Appendix 1, can be used to support the

following propositions:

(1) There is very little evidence for sexual dimorphism in tooth length in

either H. equinus or H. niger. Such sex differences as may exist are small and

there is reasonable justification for lumping measurements from both sexes in

comparing H. equinus and H. niger with each other and with the fossil samples,

which cannot be partitioned according to sex in any case. It is also relevant to

point out that the coefficients of variation (100s/x) (Appendix 1) are not sub-

stantially different for the mixed sex samples of teeth than for the single sex ones,

suggesting again that lumping the sexes for comparison with the fossil samples is

reasonable.

(2) The various teeth of H. equinus are longer than their counterparts in

H. niger, the differences being especially great for the premolars. The premolars

of H. niger variani are not significantly different in size from those of the remain-

der of the H. niger sample, but the molars are significantly longer, approaching

those of H. equinus in length.

(3) The dental samples from Elands Bay, Elandsfontein, Bloembos, Swart-

klip, Eyre’s Cave, Hawston, Lake Pleasant, and Klasies River Mouth are all

very similar in mean measurements for any given category. Relatively few

statistically significant differences can be demonstrated among these samples,

and those that do exist tend to be small. Teeth in each of the cited samples are

consistently shorter than corresponding teeth in H. equinus, the differences

being especially marked for the molars. The premolars in the various samples

tend to be significantly longer than those of H. niger, while the molars are

roughly comparable in length to those of H. niger, though small differences in

mean molar length between these samples and H. niger can be demonstrated ~

statistically in some cases.

(4) Significant differences between the means of dental samples from Nelson

Bay Cave, Oakhurst, and Uniondale on the one hand and those of the various

fossil samples considered under (3) on the other are fairly common, with Nelson

Bay, Oakhurst, and Uniondale means larger than the others in every case.

Additionally, the means of the Nelson Bay, Oakhurst, and Uniondale dental

samples differ from those discussed in (3) in being sometimes significantly larger

and sometimes significantly smaller than those of H. equinus. At the same time,

the means for both the molars and the premolars in the Nelson Bay, Oakhurst,

and Uniondale samples tend to be significantly larger than those of H. niger.

110 ANNALS OF THE SOUTH AFRICAN MUSEUM

Clearly there is justification for lumping Nelson Bay, Oakhurst, and Uniondale

together as a group distinct from the other fossil samples. The peculiar behaviour

of the Nelson Bay, Oakhurst, and Uniondale samples with respect of H. equinus

strongly suggests that they are mixed, that is, that each contains material from

more than one species, while the extent and nature of the mixture varies from

dental category to dental category within each sample. Species mixture is

particularly indicated for the Nelson Bay sample in which several dental cate-

gories exhibit relatively high coefficients of variation (Appendix 1), especially

considering the comparatively small sample sizes.

The small dental samples from Melkhoutboom and Cradock Springs behave

similarly to those from Nelson Bay, Oakhurst, and Uniondale with respect to the

other samples, both fossil and comparative, except there is no instance in which

a Melkhoutboom or Cradock Springs mean is significantly less than one for

H, equinus.

(5) The Hippotragus skull in Glasgow that Broom assigned to H. leucophaeus

is difficult to compare with the fossil samples because the only measurements

available on it are maximum lengths of the entire premolar and molar rows.

Intact molar and premolar rows are very poorly represented in the fossil samples.

Additionally, alone among the measurements presented here, those on the Glas-

gow specimen were not made by the author, but were extracted from Mohr

(1967: 62). Using these measurements, the Glasgow specimen is distinguishable

from H. equinus by its significantly smaller upper and lower premolar rows and

from H. niger by its significantly smaller upper premolar row. If the Glasgow

skull derives from H. lJeucophaeus, this difference from H. niger is difficult to

understand (a longer premolar row than in H. niger would be expected—see

conclusions below based on proposition 3 above), and it is possible that a mis-

print in Mohr is responsible. (If the upper premolar row were 45,50 mm instead

of 35,50 mm as given by Mohr, the Glasgow specimen would be indistinguish-

able from H. niger in upper premolar row length.) It must be concluded that the

present study has not clarified the specific identity of the Glasgow skull, though

on the basis of characters which are not considered here, but which may be seen

in Mohr’s photographs, I think it is highly likely the Glasgow specimen belongs

to H. niger. This conclusion has been reached independently by Gentry (pers.

comm.).

(6) Although the fossil horn cores cannot be ‘sexed’ to make them strictly

comparable to those of H. equinus and H. niger, in both of which there is

significant and substantial horn-core dimorphism, it is interesting that the

average transverse diameter of the fossil specimens is closely comparable to that

of both H. equinus and H. niger females, while their average longitudinal diameter

significantly exceeds that of the females of either species. This suggests that the

fossil males and females possessed smaller horn-core bases than the correspond-

ing sexes in either H. equinus or H. niger, while the degree of transverse com-

pression characterizing each sex in the fossil group was more comparable to that

found in the sexes of H. niger than in those of H. equinus.

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 111

CONCLUSIONS

It seems reasonable to conclude that the relatively homogeneous material

from Elands Bay, Elandsfontein, Bloembos, Swartklip, Eyre’s Cave, Hawston,

Lake Pleasant, and Klasies River Mouth derives from a single species of Hippo-

tragus which may be differentiated from H. equinus by substantially smaller

molars and premolars, and by smaller, more transversely compressed horn cores.

It may be differentiated from H. niger by its larger premolars, higher premolar

row to molar row length ratio and smaller horn cores. Since the various sites lie

in or near the area where H. /eucophaeus was encountered historically, and since

there is nothing in the contrasts with H. equinus and H. niger that is contra-

dicted by historical accounts of H. /eucophaeus, it is only logical to assume that

the species represented at the fossil sites is Hippotragus leucophaeus. The fact

that a few small, but significant differences exist among the presumed H. /euco-

phaeus samples is not surprising, considering their spread over a span of tens of

thousands of years.

The most economical explanation of the relatively heterogeneous material

from Oakhurst, Nelson Bay Cave, and Uniondale is that it results from a mixture

of H. leucophaeus and a closely related, but significantly larger species. The most

reasonable candidate for the second species is H. equinus, though if this is

accepted, the data in Appendices | and 2 imply that the now extinct southern

Cape H. equinus was significantly larger than the recent central African variety

that dominates the comparative sample. The probable distributional overlap of

H. leucophaeus and H. equinus at Uniondale, Oakhurst, and especially at Nelson

Bay, where overlap seems to have lasted for several millenniums, clearly suggests

that the two forms are separate species and not simply subspecies. As indicated

by Mohr (1967: 20-21), many nineteenth-century authors and some twentieth-

century ones have regarded H. leucophaeus as only a subspecies of H. equinus.

Interestingly, overlap in the vicinity of Nelson Bay Cave may have continued into

historic times if, as Mohr (1967: 16) reasonably suggests, an animal seen and

illustrated in 1778 by Gordon near Plettenberg Bay was a roan and not a blue

antelope.

The small samples from Melkhoutboom and Cradock Springs are most

reasonably assigned to H. equinus. Whether or not H. leucophaeus was also

represented in the vicinity of these sites must remain uncertain until larger

samples are collected.

The fossil data suggest that H. leucophaeus was both more widely distri-

buted and more numerous in the past than at time of historic contact. In the

earlier part of the Last Glacial, it occurred both east of Plettenberg Bay (at

Klasies River Mouth) and west of the Hottentots Holland Mountains (at

Swartklip and perhaps also at Elandsfontein).? During this time interval (roughly

between 70000 and 35000 B.P.), H. leucophaeus may have been the only

2 It is the geological antiquity and not the specific assignment of the material from Elands-

fontein that is problematical.

jh 2 ANNALS OF THE SOUTH AFRICAN MUSEUM

species of Hippotragus in the southern Cape (at least south of the mountains of

the Cape Folded Belt). Both well-dated, earlier Last Glacial sites have provided

fairly large Hippotragus samples with no hint of species mixture. At both sites,

the frequency of H. /eucophaeus vs. that of other taxa in the fossil assemblages is

relatively high, suggesting that H. Jeucophaeus was a fairly common antelope.

H. leucophaeus may have maintained its broader-than-historic distribution

throughout the Last Glacial, though this cannot be established at present. It can

be said, however, that it was fairly numerous near Nelson Bay Cave near the end

of the Last Glacial, while in the early Holocene, it once again occurred far out-

side its historic limits, as far west as Elands Bay and as far east as Uniondale

Shelter.

The time when the ranges of H. equinus and H. leucophaeus first overlapped

remains uncertain, but some clues may be obtained by examination of dental

samples from the different levels of Nelson Bay Cave. Although there is no

statistical technique which will separate mixed Nelson Bay samples into discrete

H. equinus and H. leucophaeus subsamples, it is possible to use an arbitrary

cut-off point to determine if the extent of mixture seems to have changed through

time. Molars are more useful than premolars in this context since they contrast

more in mean size between H. equinus and H. leucophaeus (as found unmixed at

Swartklip, Klasies River Mouth, etc.) than do premolars. Limited experimenta-

tion with the data in Appendix 1 showed that a useful arbitrary cut-off point was

the mean length for each wear category of each H. equinus molar minus the

standard deviation for that category. In Table 3, the number of molars smaller

than this arbitrary standard in each of the major culture-stratigraphic units of

Nelson Bay is presented in the left-hand column, the number of molars larger

than the standard appears on the right. For the sake of comparison, the

apparently homogeneous sample from Klasies River Mouth, considerably older

than any of the Nelson Bay samples, has been partitioned in the same manner

and included in the bottommost row of Table 3.

Table 3 clearly shows that the 11 000-8 000 year unit at Nelson Bay (Albany

culture-stratigraphic unit) is characterized by a significantly larger number of

teeth, longer than the arbitrary standard than are the other two Nelson Bay units

or the Klasies unit. These data may be used to argue strongly that H. equinus

only became prominent at Nelson Bay after 11 000 B.P. or even that it only first

appeared there at that time. More sophisticated techniques applied to larger

samples than are presently available may allow a more conclusive statement at

some future date.

11 000 B.P. was a time when not only culture but also environment was

changing at Nelson Bay Cave. Evidence from analysis of the entire fauna (Klein

1972) indicates that extensive stretches of grassland present prior to 11 000 B.P.

were shrinking, probably as a consequence of bush-forest encroachment,

perhaps in combination with drowning of much of the coastal plain by the ter-

minal Last Glacial rise in sea-level. It is entirely possible that this environmental

change was what brought H. equinus into the area, or if it was already there, led

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 113

Table 3

Comparison of the frequencies of Hippotragus molars above and below an arbitrary standard

length in the different culture-stratigraphic units of Nelson Bay and Klasies River Mouth caves.

Number of molars Number of molars

smaller than the larger than the Approx.

Culture-stratigraphic H. equinus mean minus Hi. equinus mean minus C-14 years

units one standard deviation one standard deviation BP:

from the H. equinus mean | from the H. equinus mean

i 5 000

A B

Wilton »OO/),) 4 (44%)

(Nelson Bay)

8 000

Albany 11 (20%) 43 (80%)

(Nelson Bay)

11 000

E F

Robberg 26 (74%) 9 (26%)

(Nelson Bay)

18 500

MAJOR TIME BREAK

> 30 000

G H

Middle Stone Age 105 (76%) 34 (24%)

(Klasies River Mouth)

280 000

Chi-square values

AB _ ee bf

CD — 5,027, p = 0,05-0,02

CD _

EF ~ 25,413, p < 0,001

BF ie

Ga 0,024, p = 0,90-0,80

to its increase relative to H. leucophaeus, as suggested by Table 3. If so, we have

evidence for a possible ecological contrast between H. Jeucophaeus and H.

equinus—tl ough both were probably mainly grazers, H. leucophaeus could be

inferred to have preferred somewhat more open habitats than H. equinus. The

same sort of contrast has been observed between H. niger and H. equinus (in this

case it is H. equinus that seems to prefer the slightly more open situations—

Child and Wilson 1964), though very little else is known about their ecological

differences. It is interesting that the frequency of H. leucophaeus vs. H. equinus at

Nelson Bay may have shifted again after c. 8 000 B.P. (as indicated by Table 3),

since there is independent evidence for further environmental change around

Nelson Bay at that time. Unfortunately, its precise nature remains undetermined

(Klein 1972a).

114 ANNALS OF THE SOUTH AFRICAN MUSEUM

At the time of historic contact, it seems probable that H. /eucophaeus was

already very much reduced in range and numbers from prior times and that

Europeans and European weapons only delivered the coup de grace. It is unclear

what may have led to its decline prior to the arrival of Europeans, but the

principal factor may have been habitat deterioration, following the introduction

of domestic sheep. These have now been documented as early as A.D. 400 for the

south-western Cape (Schweitzer & Scott 1973). Interestingly, H. equinus is

presently in a state of decline which is not clearly linked to human predation

(Ansell 1971: 46), but which may in fact relate in complex fashion to the intro-

duction of domestic stock over much of its range. Future research at archaeo-

logical and palaeontological sites in the southern Cape should provide data to

help explain the decline of H. leucophaeus as well as more information on its

appearance, habitat preferences, and past distribution.

SUMMARY

The blue antelope, Hippotragus leucophaeus, was encountered by early

European travellers to the Cape in a small area centring on the triangle Swellen-

dam—Caledon—Bredasdorp. It was apparently never numerous in historic times

and became extinct around A.D. 1800, before qualified scientists could make

observations on live specimens. This paper analyses relevant fossil material from

several Upper Pleistocene and Holocene localities in the southern Cape in an

attempt to resolve some outstanding questions on the taxonomic status, distri-

bution, and ecology of H. leucophaeus. It is concluded that it was a good species

(not simply a subspecies of the roan, H. equinus), that at various times in the past

it was both more numerous and more widely distributed than at time of historic

contact, and that it probably preferred somewhat more open habitats than its

close relative, H. equinus, with which it apparently overlapped throughout much

of the Holocene, at least in the vicinity of Plettenberg Bay.

ACKNOWLEDGEMENTS

Q. B. Hendey provided constant advice and encouragement for this study,

most of which was carried on in his department at the South African Museum.

M. Brooker, H. J. Deacon, J. Grindley, Q. B. Hendey, J. Parkington, R. Singer,

and J. Wymer kindly made relevant fossils available to me, while C. K. Brain,

A. W. Gentry, Q. B. Hendey, L. de la Torre, and V. J. Wilson gave me access to

important comparative collections. During the course of the study I benefited

greatly from discussions with H. J. Deacon, A. W. Gentry, and Q. B. Hendey.

K. Scott drew the illustrations of important fossil specimens. D. Eggers and

G. Johnson prepared the data for computer analysis. A. W. Gentry and

Q. B. Hendey provided helpful criticisms of a preliminary draft of the manu-

script. National Science Foundation grant GS-3013 provided the funds that

made the research possible.

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 115

REPERENCES

ANSELL, W. F. H. 1971. Order Artiodactyla. Jn: MEESTER, J. & SETZER, H. W., eds. The mam-

mals of Africa: an identification manual. Part 15. Washington, D.C.: Smithsonian Institu-

tion Press.

Broom, R. 1949. The extinct blue buck of South Africa. Nature, Lond. 164: 1097-1098.

BuTZER, K. W. (/n press.) Preliminary study of the Pleistocene sedimentary sequence at Swart-

klip, False Bay, Cape Province. New journal of uncertain title.

BUTZER, K. W. & HELGREN, D. 1972. Late Cenozoic evolution of the Cape coast between

Knysna and Cape St. Francis, South Africa. Quaternary Res. 2: 143-169.

CHILD, B. & WILSON, V. J. 1964. Observations on the ecology and behaviour of roan and sable

in three tsetse control areas. Arnoldia, (Rhodesia) 1 (16): 1-6.

Cooke, H. B. S. 1947. Some fossil hippotragine antelopes from South Africa. S. Afr. J. Sci.

63: 226-231.

DEACON, H. J. 1969. Melkhoutboom Cave, Alexandria District, Cape Province: a report on the

1967 investigation. Ann. Cape prov. Mus. 6: 141-169.

Deacon, H. J. 1972. A review of the post-Pleistocene in South Africa. S. Afr. archaeol. Soc.

Goodwin Ser. 1: 26-45.

GENTRY, A. W. 1966. Fossil Antilopini of East Africa. Bull. Br. Mus. nat. Hist. (Geol.)

12: 45-106.

GENTRY, A. W. 1970. The Bovidae (Mammalia) of the Fort Ternan fossil fauna. Fossil Vertebr.

Afr. 2: 243-323.

Goopwin, A. J. H., DRENNAN, M. R., & SCHOFIELD, J. F. 1938. Archaeology of the Oakhurst

Shelter, George. Trans. R. Soc. S. Afr. 25: 230-323.

HARPER, F. 1945. Extinct and vanishing mammals of the Old World. New York: American Com-

mittee for International Wildlife Protection.

HENDEY, Q. B. 1969. Quaternary vertebrate fossil sites in the south-western Cape Province.

S. Afr. archaeol. Bull. 24: 96-105.

HENDEY, Q. B. 1974. The Late Cenozoic Carnivora of the south-western Cape Province.

Ann. S. Afr. Mus. 63: 1-369.

HENDEY, Q. B. & HENDEY, H. 1968. New Quaternary fossil sites near Swartklip, Cape Pro-

vince. Ann. S. Afr. Mus. 52: 43-73

KLEIN, R. G. 1972a. Preliminary report on the July through September 1970 excavations at

Nelson Bay Cave, Plettenberg Bay (Cape Province, South Africa). Palaeoecology Afr.

6: 177-208.

KLEIN, R. G. 19726. The late Quaternary mammalian fauna of Nelson Bay Cave (Cape

Province, South Africa): its implications for mega-faunal extinctions and environmental

and cultural change. Quaternary Res. 2: 135-142.

KLEIN, R. G. 1974. Environment and subsistence of prehistoric man in the southern Cape

Province, South Africa. Wild Archaeol. 5: 249-285.

KLEIN, R. G. (dn press.) The mammalian fauna of the Klasies River Mouth Caves. Jn:

Wyner, J. & SINGER, R. The archaeology of Klasies River Mouth. Chicago: University of

Chicago Press.

Monrr, E. 1967. Der Blaubock, Hippotragus leucophaeus (Pallas, 1766). Eine Dokumentation.

Mammalia Depicta: 1-81. Hamburg: Paul Parey.

PARKINGTON, J. E. 1972. Seasonal mobility in the Late Stone Age. Afr. Stud. 31: 223-243.

SCHWEITZER, F. R. & Scott, K. J. 1973. Early occurrence of domestic sheep in sub-Saharan

Africa. Nature, Lond. 241: 547-548.

WELLS, L. H. 1960. Mammalian remains from Late Stone Age sites in the George—Knysna

area. S. Afr. J. Sci. 56: 306.

WELLS, L. H. 1970. A late Pleistocene faunal assemblage from Driefontein, Cradock District,

C.P. S. Afr. J. Sci. 66: 59-61.

Wyner, J. & SINGER, R. 1972. Middle Stone Age occupational settlements on the Tzitzikama

coast, eastern Cape Province, South Africa. In: Ucxo, P. J., TRINGHAM, R., & DIMBLEBY,

G. W., eds. Man, settlement and urbanism: 207-210. London: Duckworth.

Wyner, J. & SINGER, R. (Jn press.) The archaeology of Klasies River Mouth. Chicago: Uni-

versity of Chicago Press.

ANNALS OF THE SOUTH AFRICAN MUSEUM

116

Lig? =

pESs‘E a

€ _—

Of 6h =a

a I

ok OS‘Sr

6d WW Sefewej sole

dyyaong

00*TS

JUDSDA]q AYDT

1UDIADA Addu * FT

6€ a

IZ1‘7 =

(4 I

00°rs Or's¢

saispjpy = dj 4am

€°8 0s

86r'P 16L‘°Z

€ g

00°rs os‘ss

Saispjpy = dij yao Mg

I I

OSs‘sr Or‘0s

MOsSEIIH + soquaojg

L’s 9p OL

SIS‘ SZ0°7 I80°€

cS Sc £¢

60°bb 00° (4547

64H WW sefewej, Sole

‘ddsqns dadiu *

SV 8°¢ 9°

8Ir'Z E80°E Lve'l

cl S 9

SEES Or‘es ZL'7S

ié‘dH WW Sefewejz sole

‘ddsqns sadiu “YY

8"P GE vv

S8E7 T09°T O81°Z

Cr v L

09°6r Se‘6P 90°6r

ié‘ddWW seewejz sole

‘ddsqns 4adiu *H

I I

OL‘9r Or‘es

uojsMDey Avg uosjany

9°P 19

O£E'Z S00'E

Ie IT

ZO IS S0‘0S

ié‘daWW Sefewe4

snuinba *{

€°S 61

Z60°€ 976°7

OV cl

88°8S 97'09

é‘dTWW sere,

snuinba ‘HT

9°9 pL

L78°€ LSE'y

O£ L

Ip‘L¢ €T'6S

LidA WW sefewey

snuinba ‘HT

— s

I N

06th *X

SaISD] YY

soTey

rd P—zdP

Sb A

OPl‘z Ss

Ls‘9g x

sore

*ap—‘*ap

*xoSs UMOUYUN Jo suoumoeds pure ‘soyewis ‘sare, = 2% SHA SWW ‘(1X01 00} 908 ‘sol1039}e9 IeaM 9} JO UOTIIUySp

] xiIpuoddy

OJ) IBaA\ 8eT = MT “Teo WNIPO, = AAW “Teo Aled = MW ‘IeOM ON = MN 3X0} 0} Ul polopisuod sayduies snspijoddif7 SNOLIeA 34} Ul

JOJSWIIP 9109-UIOY puvL YISUS| 4100} Yded IO} (x/SOOT = A) UONIILA JO JUSIOYJO09 pue ‘(S) UONeIAOp piepur}s ‘(N)) OZIS [dues *(x) UROW ONEWIUVIIY

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 117

OT‘0s

dy 4J4OMS

ries are De

TS0'0

I I €

7£9'0 569°0 +690

MOSSYIIH soquaojg dijyj4DMs

9°T

89I'T

CED

dyyyavng

66°dH WW sopewoy

1UDIADA Jadu °F]

£6 SH WW Sopewoy

IUDIADA Aasiu *

L644 WW sofewoy

1UD1ADA dasiu * FT

SEOr

sorey

9650

Sorel

Or‘9L

SoTe

¢‘¢ SS ¢‘¢

60S‘°7 (45 a4 VELT

9¢ SG VC

98°Cr 19°Cr 0S ‘Or

464A WW Soyewey So[eIN

‘ddsqns sasiu *H

8‘ I‘€ 69

0£0°0 0700 0-00

TS 76 (KG

790 6790 079°0

64°SH' WW Sopewe{ Sole

*ddsqns dasiu *]]

I I I

00°7L SS ofl 08‘S9

MOSSDIQH Soquadjg adv) Ss aidq

6°V SP 6'V

Orr‘ ZEEE 87S‘

€9 O€ 87

09‘0L vO‘OL €vIL

i6°dH' WW Soleo, Sole]

‘ddsqns 4asiu "

a os Ae IN

oa EE Orc. 5

if I ey N

os‘ce 0¢‘6P €L‘6b xX

MOSSDIH ADDS aldq saisvpy

Ks GS Olpme 2A

969°C CELT CLS‘T s

cs (KE IT N

€9°TS os‘zs VES x

L6 HA WAN $ Ssopewoy Sole

snuinba *H

rd —zd

gc 78 IN

8£0°0 zS0°0 ocoo =o Ss

be IT 6 N

959°0 0r9°0 1990 =X

L6°dH WW Sopewoy Sole

snuinba *]]

"W—'W/"d—"d

9°7 at A

€79°T ae s

€ I N

O€‘OL O€ PL x

ulaquofspun]q Avg uosjan

CE 9°€ ip AN

6bL‘Z 99L‘7 lee 8S

CV vi cI N

CS*LL PL‘LL Wi es

L6‘dA WW Soyewos Sole

snuinba "HT

"W—'N

ANNALS OF THE SOUTH AFRICAN MUSEUM

118

6rS‘0

MOSSD]D

00°69

Avg uosjan

600°0

808°0

sais, yy

O1‘€E9

SaIsv] yy

L‘6 can

Z96‘E =

¢ A!

78 ‘OP =

sose [[e

‘ATWAN

saisvy yy

I ———

T9L‘0 ==

66°SH WW Seewe,{

1UDIADA JOSIU *

dy y.4Mg

ev a:

LSV€ =

€ oe

LE‘L9 =

6d WW soyewoy

1UDIADA Aasiu *

1] >4140MS

s}ynpe

qewoy

IUDIADA AOSIU * FT

S1‘09

synpe

Lv9‘0

SOIR

p°8

LES‘€

I7Cr

snolaadd aady) fo wing

l‘91 L‘6 rol

6869 PS8°€ Ost‘s

(Ne, Li IT

Irer 79°6E Cr 6r

sose [[e s}jnpe syjnpe

AAWN eewo, s[ePW

‘ddsqns sasiu *]]

0°8 7 A

core = s

L I N

vier 097 x

ulajuo{spun]q dog uosjan

8‘rI 6°L 2 ete"

1¢9°9 8E0°E 809°¢ s

ce 9 v N

€6°CP CE8E CO’6b xX

sose [le syjnpe s}[npe

AITWAW eewejz a[eW

snuinba *{

JOJOWILIP (ASIOASULI}) [eIO}e]-OIPa|A| —9Seq 91090-UIOH{

(7 lv ev

0£0°0 6200 0£0°0

VS VC VC

O0L‘0 80L‘0 L69°0

L6 ST WW Seewe,z

‘ddsqns dasiu *]]

Sore

aS aS (as

Osr‘€ SLE pSV'€

89 Of Le

85°S9 p9°P9 €9°99

£6SH WW Sefewe4

‘ddsqns dasiu *

SoTeI

Lv 9°P 6G = A

S00 v£0'0 rro'0—s Ss

vs (G6 Ol N

8rL‘0 SrL‘0 OrL‘o x

£é6SH' WW Soleo, soleW

snuinba "HT

eW—iW/+d—2d

= a A

== == S

I I N

0s‘99 0S*69 x

MOSSDID ulajuofspun] qT

(Oh 8°€ Oye aN

$86°C OL9°7 OVEE ies

99 €C SIT N

LL‘OL SeOL CooL &

Lé SS WW Sofewe,s sole

snuinba "7

seN—iW

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 119

as ve 61

= IZO'T 6790

I el el

O€ ZI Grer S6-TI

WI MW MA

66d WW —Snuinba +H

cv0'0

v6L‘0

SaIsv]y

9‘OI

78P°¢

ps‘Ts

Saisvy yy

¢‘¢ = 9p

LEL‘O = 99S‘0

v = Z

SE‘el = 07 ‘CI

MN MT MW

soyewo,y —snuinba “FT

I zs I

97L‘0 = 97L‘0

sose [[e s}[npe sy[npe

‘AWN 9Fewoy se

1UDIADA Aaclu *

ae a I

=e = 68°78

sose |e s}jnpe s}jnpe

‘AWN 9ewey s[eIN

1UD1ADA dadiu *

8'P = =

re9'0 zoe. =

v I as

COE! Ole 7

MA MN WT

I'v

€£0°0

1080

snoidadd aasy] fO wns

0°8 v'V s*¢

€90°0 9€0°0 0r0'0

Le (Ei) IT

88L‘0 S780 6CL‘0

sose [[e s}[npe s}jnpe

FATWA 9ews, se

‘ddsqns sasiu "HH

66 oS =a

ZIT T €L9°0 =

9 fl I

EGaL 00°ET OS ‘ZI

MW Ma MN

SoTeJA| —Snuinba "FT

9°€ =<

620°0 pee

L I

8080 980

ulajuofspubn] qT Apg uosjan

SV 8°7 l'v

6£0°0 SZ0°0 ve0'0

Ge 9 v

€68°0 6980 0£8°0

sose [[e synpe s}]npe

‘ATWN 9ewey s[eIN

snuinba "Ht

Late FL, a>

TL a>

IOJOWILIP 10119}SOd-0.10}UL/19}OWLIP [e.19}e]-OIPoJA| — 9Seq 9109-UIOH]

L'8

Ely

8L'ts

Snolaadd aadyj fo Ung

9°CZ 66 671

8E9°C1 7LL't T9L‘8

Le El IT

L6SS Ilr T1‘89

sose [[e syjnpe s}jnpe

‘AAWA 9ewey Te

‘ddsqns 4asiu *"

v'8 a

IIS‘ ae

L I

Lvs 07'S

ulajuof{spun] 7 Avg uosjan

lst L‘6 Isl

Trl‘6 LIE‘y 096°8

ce 9 v

p9‘0S Or'br 8E‘6S

sose [|e s}npe s}jnpe

“SAWN Fewey s[eIN

snuinba *]]T

ix 7, a>

ix Z, a>

JQOWIeIpP (TeUIPNUSUOT) IO1I19}sod-o19}uy — 9Seq 9109-UIOH,

ANNALS OF THE SOUTH AFRICAN MUSEUM

120

59 a SL —

966'0 — ZIZT =

L I € =

rr‘sl os‘sI Or‘9I —

MT MW Ma MN

42 ‘Aa WIA ‘ddsqns 4asiu “Hy

L‘L g‘€ r‘0l rans

ZIP‘ rEL‘0 L88‘I 066'0

61 v ¢ Z

87‘8I 0s‘61 O1'sI 06'81

AT MW Ma MN

62 ‘AAW —Snuinba *H

ae 8‘T O°ST —o

% 8070 168°T mee

I E ¢ T

Os‘TT 09°TT 79°71 Olt

MT MW M4 MN

Avg uosjan

Z¢I 9'L 9°6 =

SEL'T 8LL‘0 EZO'T r=

€ C L =z

OTT SOL p90 ee

NT MW MA MN

06 ‘Aa WW—‘ddsqns sadiu “7

S‘s = 8°9

6r8'0 = 190°T

C = c

Ors = os‘sT

MT MW Ma

soyeus,j —"ddsqns 4adiu "yy

SY €*¢ si

€b8‘0 190°T 7

v (4 I

69°81 8°61 06'r1

MI MW MA

soyewioj —snuinba *

=a a €°8Z

= ri 8I8°€

I == C

076 a Os*€el

WT MW Md

SaIsvIy

= = Is

Sy 7 €1S‘0

aaa T €

= OL‘6 El‘Ol

NT MW M4

soyewo.j —‘ddsqns sasiu *

I —

os‘sT aes

MW Ma

soe — ‘ddsqns sasiu "yy

= ‘I

== €87'0

I C

O16 os‘8I

MW Md

sore — snuinba “FH

I —

OE‘OT 4

MW MA

juvsvafd ax0T

= O'rI

as birt

= cE

oa Or'ol

MW MA

1] {340M

+a Gu

= €6l'T

I €

08°01 L9‘OI

MW MA

sojey — “ddsqns sasiu “yy

Sik

= s

— x

al At

—— Ss

a N

— x

*aP

=")

== Ss

— N

—

Lie =A

trlo 6s

Sani

rS‘OI x

= A

== Ss

—= x

121

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE

9'p Lc =< 0'r

Pelt 8LL‘0 [=e O€T'T

v c I ¢

0 4 60°67 OL‘TE 8787

NT MW Ma MN

Avg uosjan

9°¢ 87 <> =e

P87'T vOL‘o ae Ses

8 v arg ae

98°77 C6'P7 es aoe

NT MN Ma MN

i6 ‘AA WW ‘ddsqns 4asiu "7

19 69 ae a

67S‘T 976'I ce Sar

17 Ol I I

CI‘SZ GLUT OO‘TE O£°67

NT MW Md MN

ié ‘dA WW —Snuinba *H

8‘Z i Cas, LY

¢6r'0 rSs‘0 S60 ZL8°0

(6 € tS Ml

C6LI 06‘LI 0781 0S‘8I

MT MW Ma MN

dog uosjan

6€ =a 8°¢ ‘7

€06°0 a O19'T €0L‘0

v I g 9

8I°€Z 09°67 p9°LZ S6°LZ

WI MW MA MN

SaIsv]L yy

L‘9 ies = =F

9cs‘T Sm ae ae

(4 C = =a

O1'EZ $9'PZ a ie

NWT MW Md MN

soyeuio.j —‘ddsqns vasiu "FH

lie v'L as ae

Tss‘0 060°C a =

€ v I ==

€£°97 87°87 OO'TE om

WT MW MA MN

soeuse,j — snuinba °F]

9°C oo a (Sa

Lzv‘0 Sie 6160 €LS0

v ee (4 8

€S‘Ol = S161 by‘ Li

MT MW Md MN

saisvpy

= 0°8 = oo. A

ae lee = sist §s

7 C I o>) oN

Fa 0S*97 OL‘8Z LS*LZ

MI MW MA MN

di) 4140MG

6S eas = — A

6€E'T a ae caer 8

9 I a = N

8L°7Z OS*PZ a aa x

M1 MW Ma MN

soe — ‘ddsqns sasiu “7

I's L‘9 ee a A

pb7'l S78" T Te = S

Cl t we an N

LS‘vZ O1'LZ aa = x

a | MW MA MN

sole — Snuinba °F]

"ap

7 Sas = <= A

— — — — s

= I ass mee N

ane 00°LT = ee x

NT MW MA MN

juvSva]q 2Y0T

oe S‘€ ae CS. =k

sae 9950 He €r60 Ss

Ts C Set 9 N

ae 07°91 Si LG =

NWT MW MA MN

d1j{J4DM'

ANNALS OF THE SOUTH AFRICAN MUSEUM

122

€‘¢ 8°L r¢ oi

C80°T 1E€9°T €ZL‘0 Saas

97 8 € I

0S ‘OZ 00°IZ L9‘IZ 08‘IZ

MT MW Ma MN

i. ‘AA WW Snuinba -

= —_ = 91

= s = 7S70

I = = €

08‘9T = a €7‘9I

MT MW Ma MN

Avg uosjan

LAS = 78 =

€16°0 = 997'T =

6 I € a=

86'S 07ST ec‘cl =

MT MW Ma MN

66 ‘dA WW— ‘ddsqns Jadiu "A

6°€ Sis L‘9 19

1690 SEO'T L8I‘T 190°T

LZ 8 v Z

Gaal IS‘LI €9LT CT LI

MT MW Ma MN

66 °dd WW —Snuinbo “AT

S°8

8Ze‘T

€

€9°ST

9°7

690

9L‘LI

v's =

ZOL‘T ar

9 I

€S‘0Z OL‘0Z

MIN MA

soreuls.j — snuinba *

wv ae

9690 =a

E I

08‘ST O8‘ST

MN MH

SaIsvLy

ae 68

a Evel

= GC

ae COST

MW M4

soyewioy—‘ddsqns sasiu “7

CC ae

por‘o =

€ I

EV'sI 06ST

MW Ma

soeua —snuinba ‘FY

I a

07°S7Z =a

MW Md

JUDSDA] AYVT

Os‘TT

L‘v ae =

£960 a4 Fi

ol 7 I

€£‘07 C77 08°07

MT MW MA

soley — Snuinba “FY

== (‘S76 ae

5 6LE0O —

=< € —

ia €9°9T =

MT MN Ma

d14JADMS!

i <a ==

€ZL‘O a =

9 I ra

Cl‘9OT O7‘ST =

NT MW MA

soe — ‘ddsqns sasiu "FY

L‘v 6 =

8080 00L‘0 aoe

OL € I

9¢°LT OI‘sI Ost

NT MW Ma

sore — Snuinba “Hy

wet 9‘¢ eer

=e. bir a

=— G —

== Or‘'sZ pie

NT MW Ma

ulajuof{spun]q

=< A

— Ss

= N

— x

edP

a A

= S

ar N

— «x

= A

= S

‘ad N

— x

a A

——— S

= N

= x

zdP

=a A

——s Ss

a N

— x

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 123

€°8 ye — =

61ST SIT = Mies

Ol € I Seer

Ip'st €£‘6I O€ 61 —

MT MW Ma MN

i ‘da' WW — ‘ddsqns vasiu “Fy

€'¢ 6°S — 67

10c'T 697'T ze 9£9°0

SC IT I Z

$6 02 €9'IZ Or‘ZZ SIZ

MT MW Ma MN

64d WW— Snuinbo "FT

I I = re

OE'sI O16! am:

MT MIN MA MN

uojsSmMOy]

as van m 01

== te a ZIZ‘0

I a e C

00°0Z = = STIZ

MT MW Md MN

dog uosjan

09 are L‘9 a

ONT cs 8S7'T —

8 = € res

SEs == L9‘81 =

NVI MW Md MAN

i6 ‘AA WAW—‘ddsqns vasiu +

68 = ra

7S9'T ae mas

t i I

09°81 07'02 O€*6I

WI MW MA

sopewia,j—‘ddsqns sasiu *

8°¢ v9 a

bre'l Eel cs

8 € ie

9S°T7Z 07 IZ ee

MI MW MA

soeuia,{ —snuinba ‘FT

NT MW MA

IAD) S asd

I I C

09°8T Oc ‘6I S707

WI MW MA

Sasol y

€°6 =e 6€

voL‘T ar LOL‘O

€ as (4

€est = 00°81

WT MW MA

soyeuioj— ‘ddsqns vasiu ‘7

OL'8I

09°81

L‘8 a oe

68ST =a a

L I =

ce°sI OL’LI a

MI MW Md

soe — ‘“ddsqns sasiu "yy

9°€ 60 Fe

SEL‘0 9070 =

Ol v I

9r'07 87°77 Or'7Z

WI MW Mo

soley —Snuinba "FY

— I —_—

ae O€ 61 ae

WT MW MA

ulajuofspuon]q

—_— ¢ —_—

aa L0‘0@ ==

NVI MIN Md

di] 4140MS

9'P ae EF

6£8°0 at =a

¢ — —

9E°81 ras? rae

WI MIN Ma

soe — ‘ddsqns vasiu "yy

rae A

——— Ss

ae N

— x

a A

= S

== N

an x

rdP

ee A

== Ss

=i N

a x

= A

ass Ss

oa N

ars A

== S

ae N

ome x

ANNALS OF THE SOUTH AFRICAN MUSEUM

124

= 601 6€ I‘¢

me TLY‘I ssr‘0 9€9°0

I ¢ € Z

07‘ Sr‘ O8‘TT Sr‘ZI

MT MW Ma MN

SaISD] YY

= 69 v6 ZO

= b0s‘0 STII O1ET

I L Ze ZI

Oro! LOTT 96'IT 18‘ZI

MT MW Ma MN

dé ‘AA WW ‘ddsgns 4asiu “A

os TTI 9°L EL

= v79'T 6£0°T €56°0

Z 9 9] 8

09°€T Se‘el 09°€T v0'el

MT MW Ma MN

i6‘dASWW— Snuinba “H

me = a 9°9

Fe 7 =a 10S ‘T

Fae — oa t

oe L9°7Z

WI MW Ma MN

Adpg uosjan

— we 0TI =

on S 67S ‘I age

— ——_ v ——S

=a =s Sliare| ween

NTI MW M4 MN

d1j4s4oMy

— €‘¢ es eS

— L79‘0 900‘T 9790

I v cI ¢

Ol‘or 06°TT CI‘ZI BEI

NT MW MA MN

soyeus.j —“ddsqns svasiu "7

am ie = 6€

= ae L6r‘0 v160

=e I v v

== 08‘OI 08‘ZI C971

NT MN Ma MN

soyewuoy — snuinba °F]

I — — ——

OL‘LI == —— =

NT MN MA MN

ajopuoluy)

=—s = jill —

aa =e ZIZ0O aa

=; I C 4

ra os‘sT C661 Cl‘6I

NWT MW MA MN

saisoyy

=a I

oa 09°71

MW Ma

SOV] —1UD1UDA Addu “YY

76 SIT

0SO'T L8E‘T

€ el

Lert LLTI

MW MA

soe — ‘ddsqns sadiu "yy

= 8°9

= 886°0

I 9

06°€T 871

MW M4

Soe — Snuinba *H

MN Md

JAD) Sash

I —

O€'07 =

MW Md

dij y140MG

a Albee a ae

aa A a>

Dn 2, a>

KZ, a>

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 125

ie €‘p ey 69

Tell 9IL‘O LZL‘0 6LI‘I

z ¢ v €

00°91 9L‘9T €L‘OT O7‘LI

MT MW Ma MN

SaISDL YY

v'L 9°] 6S =

OLO‘T OLP‘Z £760 =

9 ST ZE I

Orr 78‘ST 08‘ST OS‘rI

M1 MW Ma MN

£6 AA WW— ‘ddsqns 4asiu +77

ES Iss o's =

IL8‘0 €88‘0 760 ee!

9 iB i, I

CGE LUEY Or‘'sI OE ‘ST

MT MW Ma MN

bd ‘SAW — Snuinba "A

oo I ———— ——,

— Ole! == =

MT MW Ma MN

uojSMDOAYL

9°8

saa |

09°rI

ZOL‘O

€

LL‘9IT

es 9°€

== €19°0

I v

O€‘LT 86°91

MW MA

d1]{J4DMS)

UL eS

080°T 708°0

9 91

77ST 68°SI

MW MA

soyews,j —*ddsqns dasiu *]T

ee ev

ae €6L‘0

I Ol

08°91 O€‘sT

MW MA

soTeulaj — snuinba *

ars I

a STEl

MW MA

soquaojg

OS‘rI

OL‘8I

00°€T

ar [

= SOL]

MW M4

SO[VIA| —IUD1ADA AadIU “Fy

OL L‘S

¢80'I SEIT

6 vl

LS‘SI v8‘SI

MW M4

sore — “ddsqns sasiu "yy

€°9 o'r

L60°I ZE8 0

€ 8

LELI SI'sI

MW Ma

Sole] — Snuinba *

MW MA

jsanyy0O

<a OCI

ee 68S'T

ar Or

=e 8Z°E1

MW MA

Avg uosjan’

Eeel

m7, OS KZ, OS

x 7, a>

Ds 7 a>

ANNALS OF THE SOUTH AFRICAN MUSEUM

126

rel gC Z'9 E€

ZOI'T C160 1860 00s‘0

8 oT I€ €

8E'oT pL‘ST p6‘SI OST

MT MW Ma MN

£6 ‘AA WW— ‘ddsqns sasiu +

o's 9‘¢ o's =

b£60 8s9‘0 Z£6'0 —

6 ¢ SZ —

bL‘st pesi IS‘8I —

MT MW Ma MN

66 AAS WW —Ssnuinba “7

I = uae

Or‘st — a As

M1 MW Ma MN

ajopuoluy)

a , = =

= 09°LT --- =

MT MW Ma MN

soquao|g

9°9 6'v 69

L70'T 68L‘0 660° I

¢ ¢ 91

REG PI‘9ol 00°91

NI MW MA

soyeuo-{ —‘ddsqns 4asiu "yy

6P = 97

968°0 ar L8r‘O

v I 6

SI'S OL‘8I 08‘8I

MT MW MA

soyeuuoj — snuinba *

MT MW MA

jsanyyv0O

== Is <=

cae 9¢9°0 ae

~ G GC

= CILI O7‘LI

NT MW MA

ulajuofspuvn] 7

09°91

OL‘rI

MIT

09°LI

OL‘6I

8°9 9°¢

L90°1 8180

8 el

19°C 8L‘SI

MW MA

soe — ‘ddsqns sasiu “yy

Le 8°9

Iss‘0 Isv'l

€ 9

LOLI BEI

MW MA

Soe —Ssnuinba *y

MW MA

WOo0gjnoyxjayW

I —

SSSI se

MN MA

u0JSMDET

6€ €°8

60L°0 SrP'T

v 81

E781 OSs‘LI

MIN Md

Avg uosjany

ee AK

pseo s

c ON

COs) ox

= ys

— Ss

— x

SIN

—_— Ss

— x

—— Ss

— x

= A

—— S

a= N

— x

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 127

0°01 8°¢ 6's a

6IL'I LSV‘I 9ST I ae

81 8P v =

O7'LI ZITO SEI ae

MT MW Ma MN

0. ‘do WW— ‘ddsgns sasiu "yy

GL 1°9 8°S L‘ol

S8E'l 66£'1 SSE] 9PS‘7

Ol Sv II (6

97°61 00°€Z So'E7 O8°€7

WI MW MA MN

6 AWN — Snuinba “H

— I —— ost

= SL‘LI =i =e

WI MW Md MN

soquaojg

al ae v7 er

MGW = ZrO ee

(4 C S I

ST‘Ol 69°91 Cou) OL‘9OT

WI MINN Md MN

SAISD]M

68 9°¢ aE =

SLP'l ZEl'T = =

8 (a6 | =

cS‘9Ol vl‘O7 06°07 =

MT MW Ma MN

soreura,y —“ddsqns dasiu "7

67 6S Gs =

v6S‘0 6LEI €06'T =

¢ Tl € =

Pl‘O7 OREO EZ ET =e

MI MW Ma MN

soyewio-| — snuinba "FT

— I — oneeeen

= 09°91 7 =

WI MW MA MN

IAD) 8 asd

I I I am

00'LT 09°81 09°LT =

NVI MW Md MAN

u1ajuofspun] 7]

oi ST S‘v as

=| €87'0 8£8°0 ae

— 08°81 cr'st ton

MT MW Ma MN

d1]{14 OMS

CO ‘9 S‘L

1881 SETI 979‘ |

8 €Z Z

€6‘LI €1‘O7Z 6607

MT MW Ma

soe] —‘ddsqns sasiu "yy

Tne 8°9 ¢‘9

— sh Z9E‘T

I 81 L

O7‘LI 9L‘°7 OL‘EZ

MT MW Ma

sole || — snuinba *

— I —e

Ta 08°91 ca

WI MW MA

uojJSMDYy

se ae v'L

ee ae viv'l

I I (4

Or'or OL‘LI 02°61

NWT MW Ma

Avg uosjan’

— I _—

a 68°91 rte

WI MW MA

SO[VIJA| —JUDIADA Aasiu "Fy

ix Z, a>

oc?” al Lc ha

WZ, a>

ANNALS OF THE SOUTH AFRICAN MUSEUM

128

8°7

pL9‘0

ZL*EZ

L‘9

SsvT

SL‘8I

vv 6€

9L0°T IZ0'T

ve cl

b9'rZ 81°97

MW Ma

62 ‘Aa WW —Snuinba "Hq

MW MA

uOJSMDFT

MW MA

soquaojg

8°¢ 79

O61‘T evel

el v

8E°07 €8°I7Z

MW MA

saisoy yy

91°97

769°0

76°C

x6 v'P 8‘p

1S9‘0 ZLO'T ZL7'T

tb IT C

L8‘€Z 79° 0£°97Z

MT MIN MA

soyeuls4 — snuinba ‘HH

O1‘€EZ

MT MIN MA

JUDSDA] AYVT

MT MW MA

ajopuoluy

9°9 GE 60

SLIT 099°0O 7170

v Sg (4

08*LT OL‘0Z $9°7Z

MT MW MA

u1ajuofspun, 7

— 6: € —

ee LLL‘O es

os 9 I

= O17 Or'€eZ

MT MW Ma

dyy.aons

s‘€

L06‘0

€

€9°STZ

06°¢Z

Or'sz

08°77

07‘9T

L‘v 8‘v ES

IOI‘ £97 61L‘O

Ol Li; ¢

C8‘r7 9797 79°97

MAW Ma MN

Sole — snuinba *

“AW

I en —_—_—

09°91 <— mae

MW MA MN

IAD) 8 asdq

MW Ma MN

jsAnyyDE

69 v9 v‘€

96r'T €79'T 8LL‘O

€l 9 C

OrIZ €EST CL‘7Z

MW MA MN

dog uosjany

MW MA MN

SOT —1uv1AddA AasIU *

oA a>

ee KZ o> a faa

a a a>

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 129

= 06‘7Z 0S ‘67 —

MT MW Ma MN

JSANYYDO

—— I — —

= ¢s‘tZ a —

MT MW Ma MN

soquao|g

= VO! 79 CL

= 88L‘Z vir‘ 8L9‘T

I €T [i v

00‘0Z CIEGe 96°77 EEC

MT MW Ma MN

Salsv] yy

6€ g‘¢ 9‘9 —

69L‘0 €97'T Crs‘T a

¢ a4 61 a

ZL‘6I PL‘IZ ZEET a

M1 MW Ma MN

£6 ‘Aa WN — ‘ddsqns sasiu +7

€‘p

9r8'0

€8°61

038"6! a

MW MA

IAD) S AAA

(5 Ge

€6r'0 LOL‘O

v (4

SO'IZ O€'7Z

MW MA

ulajuof{spun] 7

¢ —

p9'1Z =

MW Ma

dyyr40Mgy

8°¢ 79

Ov7'T 97P'T

91 Ol

IS*IZ 16°C

MW Md

soyewio,j— ‘ddsqns vasiu "Hy

ZO7'T

SP'bT

— I —

cz $8°SZ 7E

AVI MW Md

wmooqginoyyjaW

I — —

00°07 mae co

WI MIN Md

uojJSMDAT

Pk Gu ETI

re TS9°T 160°€

ae 8 ¢

a 86°77 Z'SZ

WI MW Md

Avg uosjan

— I —_—

GE 66°€7 cea

MT MW Md

SOR —1uv1iva sAadiu *H

oer 0's 69

age TIT 629'T

I €¢ 6

O€ 61 10° LL*€7Z

MI MIN Md

soyeW — ‘ddsqns dasiu "yy

sienna +3

7 9l

TEO'P

$8°PZ

Ty A xz a> xz a> x7 a>

ix Z, a>

ANNALS OF THE SOUTH AFRICAN MUSEUM

130

= 65°67 = ==

MT MW MA MN

soquaolg

97 Lv 76 ¢‘0

SLL‘O 9LET Lie Ivl‘O

(é 9 v Z

CI‘OE Te°8Z 00°PZ 09°97

MT MW Ma MN

SAISD]Y

9°9 79 lef g‘€

€Z0°7Z TREE II PSI‘Z 190°1

v €Z €€ G

€6°0€ P1°67 V8°LZ CI‘8Z

MT MW Ma MN

ié°SA WW ‘ddsgns 4asiu "7

L‘v L‘v LS 07

619'T 6LY'T 60L‘T Z09‘0

9 SI IZ €

09‘PE LYTE 7Z‘OE 95°67

MT MW Ma MN

66‘dad WW — Snuinba +7

=e Cie v9

Pest L790 89L‘T

B= v C

ae OI'8Z CL‘LTZ

MT MW MA

u1ajuofspun]q

—_ G16 —_

= 9790 ars

_— ¢ ed

a 8r°8Z se

MT MW Ma

di] Y1IDMS)

9°9 0°9 9'P

EOC 96L'‘T v67 I

v 6 91

€6 0€ L6°87Z 70°87

NT MW MA

soyeua.j —‘ddsqns vadiu "YT

b'v eV JPL

66b'T LEI EST"?

€ ¢ y)

(SS 06 TE ETOE

NT MW MA

Soyeulo-j — snuinba *

00°CE

OL‘9€

I —_—

soe —snuinba "yy

a an SAL FS. See RS ee SE Maa

eae a>

06°LZ =a =

MW MA MN

IADD 8 add

TOI 69 = A

880°€ 6L0°7 =e s

IT ¢ I N

ve‘OE 96°67 OS Te. x

MW Ma MN

Avg uosjan

a Pe, = A

— — — Ss

I = im N

OS‘*ZE — = x

MW M43 MN

STV —1Uv1ADA Aas “Fy

L‘9 €Ol = A

CL6T 9S87 Tr Ss

Ol OI if N

€v'67 L9‘LZ 06°87 xX

MW MA MN

soe — ‘ddsqns 4asiu “y

€°9 oS = A

C861 99° — S

¢ 8 I N

Cele 67 OE O7‘OE xX

MW MA MN

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 131

O19 — _

MT MW Ma

uoJSMOYT

pL = ITI

1OE‘T — Z68‘T

£ = €

LOLI — OL‘9T

MT MW Ma

SaIsv]y

9°8 99 9‘L

SETI -L60 6LI‘I

6 91 I€

Lev p8‘rl ers

MT MW Ma

6 AA WAN — ‘ddsqns vadiu “7

€‘Ol L‘9 OL

ZS8‘T Iv‘ 700'T

Ol €I ZE

v6'LI Allah v69I

MT MW Ma

62 ‘Ad WW —snuinba "Hy

06°€1

I 7 I

06°91 Te O€‘9T

NT MIN Ma

ulajuofspuvn] 7

— GC —

ral SO'LI as

WI MIN Md

dij JADMs!

°8 Teg 8°8

CoG Osr‘d Sel

9 v cl

€9'rl SSSI 9E°CT

WT MINN Ma

soyeus,j —‘*ddsqns dasiu *

0°6 9°6 6S

L6S‘T OL9‘T 866.0

v Vv el

78°LI 8e‘LT L8‘9I

WWI MIN Md

soreuwlsa,j —snuinba * 7]

I — a

07'81 a a

NI MW MH

IAD) S add

0S 9 78

7080 vill OIr'l

£ £ 8

€1‘Ol O¢'sI 6I‘LI

MT MW MA

Avg uosjan

—a I —_—

oe 08‘rI =

WT MW Ma

SORIA —1UD1ADA 4ad1U * FY

ve s‘¢ VL

66r'0 6180 brit

C 8 (SL

cS‘vl €6 v1 SPs

MI MW Ma

soe — ‘ddsqns sasiu "7

a 8° GE

a 00s‘0 8811

I c L

09°LI OL‘LT Lr‘ol

WI MW MA

soyey| —Snuinba * Fy

x7 a> Ly xz, OS

IK Z, a>

ANNALS OF THE SOUTH AFRICAN MUSEUM

132

I — — er

06ST ae a ==

MT MW Ma MN

uoJSMDT

Ge == €‘8 =

680 = Z6E'T =

Z Z ¢ I

Or‘sT 69‘9T 89°OT OL‘ST

MI MW Ma MN

SAISDLY

v9 1c GS gb

0760 69L‘0 86L‘0 669‘0

6 61 T€ v

6b 80ST 87ST Ser

MT MW Ma MN

ié ‘AA WW ‘ddsqns sad “7

¢‘¢ 68 9p 9°7

8960 OLr‘T I8Z‘0 9Er'0

Ol ZI €€ €

69‘LT O¢‘LI CHET Os‘9T

MI MW Ma MN

£6 ‘dad WW —Snuinba *H

I — —

OL‘9T = a

MVT MIN Md

ajopuoluy)

I a I

O7'LI aa 06°91

WI MW Md

ulajuofspun] 7

I I rae

OL‘LI 08°L1 =

NVT MW MA

d1]414OUS)

I°9 6? 6€

8L8°0 TEL‘0 p850

E S vr

6c v1 00'sT 8I‘sT

WIT MIN Md

soyewoj—‘ddsqns 4adiu "HY

19 8°¢ (ais

blot Lv0'l 9680

9 v el

ZS*LI S6LT 8E°LT

MT MW MA

soyeule-y — snuinba "FT

OL‘ST

O€*LI

MW Ma

IAD) S add

= OTT

es 768

if v

00°61 O€‘LI

MW MA

Apg uosjan

i —-

OL‘ST os

MW Ma

SOV —1UD1DA Aadiu “YY

gC L‘9

7880 6720'T

Or vl

87ST pest

MW Ma

soe — ‘ddsqns 4asiu “y

EE 9°T

Lss‘0 €97°0

€ L

00°LI 9L‘9T

MW Ma

soley — Ssnuinba *H

xz a>

fare A

== s

ieee!

OL‘OT x

=< A

= s

sae N

<= x

Co A

p6l'0 =S

EON

OS‘pI x

s6v0—S

c WN

so'OT Xx

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 133

ZIZ0O

Cov

Or'sI = —

NTI MW Ma

ajopuoluy)

(oale L‘LI on

ZOr‘T OPS ‘7 PEO

t @ OI

LEe‘sI Or‘vl 97 VI

NT MW Ma

saisoyy

0°6 09 9‘¢

98‘ £980 7780

Il 91 OF

CCl 9r‘rl ZL‘¢l

NT MW M4

o6 AA WW — ‘ddsgns sas +7

OL Or Gc

€60'T 1L9'0 868'0

cL Ol €€

IL‘ST 86°9T 8791

MT MW Ma

62 dA‘ WW —Snuinba ‘Hq

ran I I

aa or‘st 00‘ST

MI MW Md

dog spuojq

GIGI ange 8°8

WAL <= pret

C =F C

Olor a C7SI

NT MW Ma

dj {140M

6'8 ‘9 LS

crit 0L8‘0 TE€s‘0

9 ¢ vl

08°CI Tvl Srv

NT MW Ma

soeulo,j —‘ddsqns dasiu "7

8°9 60 OL

Z80'T €csT‘O IZU‘T

9 € €l

Cast €8OT 90°91

NT MN Ma

so[euus,j — snuinba °F]

OS‘LI

9°6

Eve'l

€

€0°r1

00°9T

sojey| —Snuinba ‘HH

MW Ma

IAD) S add

ar 8°L

ae C6l'I

if IT

Os‘El SES

MW MA

dog uosjan

if —

6O'rl =

MW Md

SORIA —1Uv1dvA dasiu *

69 ¢‘S

6¢6°0 vIs‘0

6 el

99‘bI L8‘rI

MW Md

soe —“ddsqns dasiu "7

7S 97

0060 vIr‘0

v 9

cL C6 ST

MW MH

7 OS PN Oe oe, ie

We Pe a>

ANNALS OF THE SOUTH AFRICAN MUSEUM

134

— I a —

— OL‘6I a =

MT MW Md MN

u0JSMDY

L‘6 saad = =

8881 as = =

€ I = I

€S‘6l 08°IZ a OSz

MT MW MA MN

IAD) S add

89 99 8°7 9‘L

peel L6E‘T €69°0 89LT

8 8 9 Z

19°6I 9717 LOS SE°ET

MT MW MA MN

SAISD]

9°8 79 9 =

p9c‘l 7E'T SIV ae

IZ Ns € =a

v7'8I Iv‘ IZ ELEZ =

MT MW Ma MN

62 °dA WW ‘ddsqns 4asiu “Hy

eL g‘¢ ov =

L8E‘T bir! SII‘T =

IZ €9 Ol I

bro! 9p 90°SZ O€*E7

MT MW MA MN

62 ‘SA WIN — Snuinba “Hq

I —_— —

OS‘EZ — es

NI MN MA

wmooqginoyyjaW

re = ve

a = 8LL‘0

a: I G

7 O€ IZ CI‘EZ

NT MW MA

ulajuofspun] 7

19 69 =

Pre’ 99¢'T =

C C I

C8‘IZ 0S‘7Z 00°€Z

NT MN Ma

d1j4jADMgy

9°L CL i

ZIET 809‘T =~

Ol ce I

CI‘LI ELIZ 00°72

NT MW MA

soyeus-4—‘ddsqns 4asiu “HH

vs VG 0°6

Cro T 68ET 9€T77

Or 0c €

C761 (46 ZEST

MT MW MA

soyeus.j —snuinba *

Ipl‘l

LE“61

8°L

LSs‘T

€

€8°6]

p's oa

rir‘ =a

(6 —

0797 erm

MN MA

ajopuoiuy

aA 9°8

S67 PeEo7

Ol ¢

LISS CS-EC

MAW MA

dog uosjan

MW M4

SOV —1UD1ADA Aadiu “FY

¢‘9 sy

aan ae

6 if

8S‘IZ OS‘IZ

MW MA

sae — ‘ddsqns sasiu "Hy

€‘¢ 60

TEE T 8770

ST ¢

€6°PT ZEST

MW Ma

Sale| — Snuinba “Hy

9°7

0s9°0

99°bZ

Ae A ae Rae 2

ne, a>

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 135

I a I =F

09‘b7Z a 06‘PZ se

NT MW Ma MN

uoogInoYyy/aW

= Xe 77 9‘¢

= 88s‘0 C6r‘0 68L‘0

I II Z ¢

OS‘EZ C8‘IZ 69°77 81°77

MT MW Ma MN

SA1SD]

8‘b v9 ep =

LIO'! 10S‘T 090°I =

¢ cs €] ==

O¢ IZ I€‘€7% 19‘p7Z ==

NT MW Ma MN

46 ‘AH WW ‘ddsqns sasiu "yy

97 79 Ww €€

1190 6LS‘I C8E‘l LI38‘0

8 7S ST IT

69°€Z SECT 99‘CZ 66 ‘PZ

MT MW Ma MN

62 “AAW — Snuinba “7

— nee O'V 9T

ae = 6160 €SeO

ei I 6 4

= 00°€Z 68°C CSC

NT MW Ma MN

ulajuofspun] 7

Bas 9‘Z pean ae

i 99S‘0 =n Sm

i OL‘IZ OE°S7 a

NT MW MH MN

dt] y140Mgy

8° ¢°9 8°9 a

LIO'T 80S ‘T 9P9'T 7

¢ CC v 7

O€'TZ LO“EZ €€PT =

NT MW MA MN

soeuis,j —‘ddsqns 4asiu *

WG v9 0'8 Gl

LLS‘0 O£9'T 970°C 90€°0

S 81 ¢ €

96°€7 6£°S7 v7'S7 L7‘ST7

NT MW MH MN

soyeulo,j —snuinba "FT

OS ‘PZ

— I I

en O1'LZ 00°82

MINN Md MAAN

Ssuldds YIoposD

= 9°¢ SIT

= S871 878°C

me IL C

ai €8°7Z OS"bz

MW M4 MN

Avg uosjany

I —_— i

00°92 = i

MW MA MN

SO[VJAI —1UD1ADA AadIU *

€°9 Ie ae

68h I 78L‘0 aoe

ve 6 —

CUNS6 EL‘PZ oe

MW MA MN

soe — ‘ddsqns 4asiu "yy

lv ev i

690°T Let SSI'T

IT 8 €

16°S7Z T19Z L9°S7Z

MW MA MN

sore —snuinba *

7g AS i 7, A> i A

LiL a>

ANNALS OF THE SOUTH AFRICAN MUSEUM

136

if —_ — —_—

Ol'P7Z = = Sa

NT MW Ma MN

IAD) S add

19 9€ GE CT

trl L8L‘0 Cé6r'l ¢L70

v L 6 v

SI“€EZ v6 17 vL‘07 88°8T

NT MW MA MN

SAISD]

0°8 61 8°¢ I‘€

9r6'I evil €OE‘T 10L‘O

¢ VE 97 ¢

9E‘P7Z LS‘€7 Lv‘tZ SCC

MT MW Ma MN

62 “dd WIN ‘ddsqns vaciu “77

UY v'YV 6P €°9

690'T 190‘T 1ON‘T 7OV'T

9 pe 17 9

8°SZ OL‘b7 SLT LWSe

MT MW Ma MN

62 ‘dd WA—Snuinba "H

— 69 — =

= S8r'T om a,

Te SL‘7Z OS‘IZ —,

WI MW MA MN

ulajuofspun] 7

a I I I

<3 O€°7 00°TZ 00°€Z

WT MW MA MN

d1]41404S

re €*¢ GL v'P

os 8S7'T 76ST 066°0

aie IT el (4

= $9°€7Z 61°27 Or‘7Z

NWT MW Ma MN

soyewo,y —‘ddsqns 4adiu “7

ve TP 9° zz

€88°0 9€0'T ZLS‘0 =

v cl v ce

O1'9Z 68°72 07°77 =

WT MW MA MN

soyeuls-j —Snuinba

8°€

196°0

€

LI'Sz

AVI

O'€ 69 0°7

6£L‘0 Z9P‘T C6r‘0

¢ 9 C

79‘bT 8717 CL‘¥7Z

MUN Md MN

dog uosjan

I —— ae

C9°LZ == a

MW Ma MN

Soe —1ud1uDA 4adiu “HY

- o¢ (a Cia

Z8I‘T v6L‘0 €0S‘0

81 ol €

6S ‘EZ OL*7Z €0°€Z

MW Ma MN

sayeW— ‘ddsqns 4asiu “yy

g‘¢ pS S‘€

Trl 9IZ‘T LI6‘0

9 L €

LO‘SZ IL‘ 00‘FZ

MW Ma MN

sore —Ssnuinba *y

Ng: ee eZ, ee a

Ke a>

(UH{ pure dH JOJ sodv [[ev) JO}OWIeIp [eJO}e]-OIPo S00 I€ 07‘Z AUH] < off

SHSVd AWOO-NYOH 1000 > 08 87'P UH < 9H

$00‘0 4 €S‘€l AUH < WU "WwW

Z0‘0-S0'0 I OS‘€Z MDD < WU rq—*q

100°0 > ps (Si PS UH < WU 9Z0‘0 fs Ol'p DA < MZ

610°0 ps 7r'7Z oH < WU ZZ0'0 € cep Lad < Aupy

rr0'o €¢ 90°Z UH < MZ L00‘0 re LO‘S MZ < AUH

¢10°0 bs Sr‘Z AU < UH] 1z0‘O €9 9¢'Z UH < AUT

1000 > €¢ 66° MDD < UH r100 SP vS‘Z ad < 9H

100°0 > €S €9°¢ MOD < 9H 100‘0 > 9p 6b'P Lid < °H

100°0 > vs ZO'V AUH < 9H 1000 > tb €Z'P 94g < oH

100°0 > 901 69°L UH < 9H 1000 > Lv BCP MZ < °H

sW iW 1000 > 901 7'TI UH < °H

rd—sd "WW

L10‘O 89 Sr‘Z UH < AU 1Z0‘O €¢ ere AUH < €Wa

8Z0'0 6S S77 4d4duH < WWUH 100°0 (AS ZS‘€ UH < aWad

100‘0 99 19°€ Wu < °H 100‘0 IS LOE UH < OAN

1000 > ZEl 0£6 UH < 9H 100‘0 €¢ Ere UH < MZ

es — WN Z0‘0-S0'0 O€ (15.06 MDD < °H

1100 LS Z9‘7 UH < Wu S00‘0 O€ 10°€ Wu < 9H

1000 > pS 60°F MDD < UH Z00‘0 I€ 67'€ AUH < 9H

1000 > S 6¢°9 MDD < °H 100‘0 > 18 sr‘ZI UH < 9H

Z00°0 cs I€‘€ AUH] < 9H ‘dd

1000 > 601 Z8‘El UH < 9H 1000 > 0S 99°¢ UH < °H

rd—sd b£0'0 Ob 61‘Z Wu < 9H

$00‘ Z Or‘ AUH < €Wa rdP—zdP

Z00‘0 IS ZE‘E UH < aNd 1€0‘0 El (G6 UH < Wu

100‘0 > AS €8‘€ UH < MZ 100‘0 > SI erp UH < MZ

"IN" 100‘0 > Or $s‘9 tH < °H

vy 2 dp—*dp

d Jp } ees d Jp 1

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 137

‘soeuUlay = 4H ‘soe = WW ‘Mosse[D = MOD ‘s3ulidg yooprig = So ‘woo noYyYpPW = qHW ‘IsinyyeoO

= MVO ‘epuoiuy, = CNN “juesesd 94eT = dT ‘oAvD S,91Kq = OW ‘UO}SMeH = NIH ‘soquioolg = qd “UlojyUOJspurlg = L4q ‘oAvD Avg

UOSJON = OMAN ‘YINOP[ JOARY soisepy = WU “dipeng = MZ ‘lasiu FY IOyIO = UP “uvIUDA sasiu "+E = AUP ‘Snuinba "FH = oF ‘JOA, oUROY

-ludIS = d “WOpselJ JO sooIsOp = Jp ‘,) SjUSPNjG, JO onTeA = } *AIOS9}"O Aq ‘jSIXD SOOUDIOYIP UROL JUROYIUTIS AT[EOIYSIZeIS YOIYM IOJ sojduues Jo ys]

7 xIpuoddy

RRR hoy

ANNALS OF THE SOUTH AFRICAN MUSEUM

138

100°0 >

600°0

1000 >

100°0 >

110°0

1€0°0

1Z0°0

100°0 >

€00°0

900°0

100°0 >

100°0

uoIsn[900

Z10°0

Z00°0

UH < OWN

MZ < 9H

UH < 9H

MA‘dP

Waa < 9H

MZ < 9H

MN‘dP

WW¥H < 4H4H

AUH < LH

WWHYH < ON + LH + WAM

WW#H < LHd

WWEH < Want

OdN < 4dH°H

H4AUH < 4H°H

OPN + Was + LAW < 4H°H

LAd < dHd°H

Wad < d49°H

WWUH < 4dH°H

WIWYUH < NW°eH

(AUT pue ‘UP ‘oH JOJ

UI *JAJ) JO}OUIeIP JOTIN}SOd-IO1INUL/IO}OWILIP [e19}e[-OIPS|

OL‘

CEG

vS‘T

69°€

d4dUH < O€N + LAA + WU

H4°H < O€N + L4H + WU

H4AUH < LHH

H4A°H < LHA

HH°H < Wat

LH + O€N + Waa < WANUH

L4H < WANUH

Waa < WNYH

H4°H < WANYH

HAUH < WWYH

HAUH < WW°H

HA°H < WW°H

(AU}{ pure ‘UP ‘OH 1OJ UOISNID90 UI *PJ) JO}OWIP IOLI9}sod-o19}Uy

b£0'0 fl) O£°7 A4°H < Lda + OPN + WU

€70°0 IT 79°7 AH°H < LH

6r0'0 CC 80°Z HH0H < Loa

100°0 > E 7s‘9 LH < Ady

100°0 S 6€°9 Wu < AuH

100°0 > CC SO’ JHd + ON + WU < WANYH

500°0 vl 6C'E Wado < WWE

100°0 > cI 6L'V SJ4°H < WAYH

100°0 > 9¢ SL‘S AAUH < WWYH

€10°0 cl 96°C Lid + O€N + Waa < WWPeH

9€0'0 ft 6S‘ Wan < WW°H

700°0 8 96°€ H4H°H < WW9H

100°0 6I SOV HSUH < WW°H

d JP }

(AU, pue ‘UP ‘OH OJ UOISN{IDO UI *JAJ) JOJOUIVIP [e19}e]-OIPI|]

900°0 OV £672 Lid < 9H

€00°0 8€ 8I'€ Wun < °H

¢00°0 Le L67 uH < AUH

100°0 > OL 87S UH < °H

1000 > Of 98°P WWHH < dsHUuH

d JP }

(UH

pue oF IOjJ sose [[e) JO}OWIeIP IO1I9}SOd-019}UL/IO}OWIP [v19}v[-OIPs|]

100°0 > Se 76 9H < AUH

700°0 LE vee UH < AUH

Sr0'0 OL v0°7 °H < UH

100°0 > O€ Or'9 HAUH < WWUH

Z00°0 6l pS°¢ HA°H < WW°H

d Jp }

(UZ pue dH JOJ sose ][e) JoJOWIeIP JOLI9}sod-o19}Uy

100°0 SE (Ms 9H < AUH

100°0 > Ot 90°S 4HAUH < WAYH

100°0 vl SSE 4dA°H < WW9H

d JP }

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 139

09"r

LL‘@

Or'€

6S'P

09°7

90°S

bre

97°

UH < 9H

MN'd

UH < dHN

uH < GNN

UH < ON

9H < O€N

UH < 92H

MT'd

N.LH < O€N

MW'd

UH < MZ

UH < OdN

AUH < 9H

OPN < 9H

MZ < 9H

Wud < 9H

UH < 9H

Maid

UH < WU

UH < 9H

MW'd

UH < DOAN

H4H9H < WW°H

Wux < °H

UH < 9H

Md‘d

WWOH < 44°H

GNN < 9H

UH < 9H

M\TrdP

Wax < 9H

UH < 9H

MWdP

Wu < 9H

MNrdP

UH < 9H

M'TedP

UH < 9H

MascdP

UH < 9H

MW T2dP

uH < 9H

Wun < 9H

MWedP

Wan < OUN

MNedP

HH9H < WW9H

Wu < 9H

UH < 9H

MT'dP

UH < OUN

UH < Wada

UH < 9H

MW'dP

Wun < ON

UH < OdUN

Wua < 9H

UH < 9H

MT'dP

OdN < 9H

MZ < 9H

UH < 9H

MW'*dP

MZ < OAUN

Wan < COUN

Waa < 9H

MN‘dP

dT < ON

UH < 9H

MW‘dP

ANNALS OF THE SOUTH AFRICAN MUSEUM

140

aY YY

~OMTNONCO ST

Lid < Auy

UH < AUR]

LAd < aNd

UH < OUN

AUH < €HWN

OdN < dHWN

MZ < €qHN

LAW < d#HwW

L4H < AVO

LAd < MZ

CH < Want

OdN < 9H

OW < °H

aWd < 9H

Lid < 9H

Wu < 9H

MZ < 9H

UH < °H

MN WN

Waa < AVO

UH < AVO

°H < AVO

L4H < °H

Wada < °H

UH < 9H

MAN

LH < °H

Waa < °H

MN'W

UH < OUN

UH < Wat

uH{ < AUH

WWP°H < 44H°H

UH < 9H

MT'W

UH < OFAN

ON < 92H

MZ < 9H

Lid < 9H

Wun < 2H

UH < 9H

MN'W

Wu < OdN

UH < OFAN

2H < O4N

Wu < °H

uHO< oH

MA'W

Waa < GNN

Wunm < MZ

MN'W

OAN < 92H

Wun < °H

UH < 9H

MT'd

UH < €Wd

uH < 144

UH < MZ

Wun < MZ

AUH < MZ

Wux < ?H

AUH < 9H

UH < 9H

MW'd

Ws < OPN

UH < OdN

UH < Was

UH < MZ

Wum < °H

UH < 9H

Ma’d

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 141

Wud < °H

OdN < 9H

AUH < 9H

UH < 9H

MWid

Want < O€N

OdN < 9H

Waa < °H

UH < 9H

Maid

UH < GNM

UH < Od

UH < LH

UH < MZ

UH < OUN

Wun < 9H

UH < 9H

MWTed

UH < Wad

UH < OUN

UH < MZ

UH < 9H

MWed

UH < Wun

UH < OYUN

UH <°H

Med

UH < °H

MNed

UH < Od

UH < OUN

UH < Wan

UH <°H

MWTed

Coe i |

ANN

UH < O€N

AUH < MZ

uH < MZ

AUF] < 9H

UH < 9H

MWcd

UH < OdUN

UH < °H

Madd

Wut < 9H

uH < 9H

MNT W

Waa < AUH

LH < AUH

MZ < AUH

Od < 9H

LHd < 9H

Wu < 9H

MZ < 9H

UH < 9H

MAN'W

UH < OFN

Want < OUN

Waa < 4H

Wat < 9H

UH < 9H

MAW

Wun < O€N

Wut < 9H

MN‘ W

NLH < 9H

Wu < 9H

UH < 9H

MTW

ANNALS OF THE SOUTH AFRICAN MUSEUM

142

Waa < 4H

OUN < 9H

Wuna < 9H

MAaeW

UH < O€N

Wau < OUN

Wart < MZ

Waa < 4H

Wao < 9H

MNEW

UH < dHW

UH < OFAN

UH < 9H

WIEN

MZ < AUH

up < AU

Wad < 4H

Wud < °H

MZ < 9H

UH < 9H

MANN

UH < SO

OdN < SO

LAW < UH

OdN < UH

Wada < 4H

LAW < 9H

OdN < 9H

Wada < 9H

UH < 9H

MAW

Lada < S$)

Waa < SO

Lod < 9H

Wan < 9H

MNZWN

Wun < dqHW

UH < FqHW

°H < dHW

4uH < Wan

Wan < OUN

UH < OdUN

°H < OUN

UH < MZ

UHR < AUR

H4HUH < WWUH

UH < 9H

MTW

Wax < GNM

UH < CNN

Waa < OUN

UH < OFUN

Did = eH

NIH < °H

Wd < 9H

UH < 9H

MWIW

IH < °H

UH < 9H

Wan < 9H

MAW

°9H < GNM

UH < GNM

9H < OUN

UH < Od€N

UH < Wan

UH < MZ

UH < 9H

NM Tid

TAXONOMIC STATUS, DISTRIBUTION, ECOLOGY OF THE BLUE ANTELOPE 143

Gey

Mo)

tTeone-Nnot

log)

MZ < OUN

Wad < O&N

LHH < Au

Wan < AUH

ON < AUH

MZ < AUH

UH < AUH{

Wu < UH

AU] < 9H

MZ < °H

Lad < 9H

Wud < °H

UH < 9H

AVG

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FIscHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FiscHer, P.-H., DUvAL, M. & RarFry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konan, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 1960b. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und -

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

Richard G. Klein

ON THE TAXONOMIC STATUS,

DISTRIBUTION AND ECOLOGY OF

THE BLUE ANTELOPE, HIPPOTRAGUS

LEUCOPHAEUS (PALLAS, 1766)

JULY 1974

TH E OUTH

CAPE TOWN.

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 ~~ Band

July 1974 Julie

Part 5 Deel

A NEW GENUS AND SPECIES OF PLIOCENE

BOSELAPHINE (BOVIDAE, MAMMALIA)

FROM SOUTH AFRICA

Ae Wes GE NGERY,

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 21, 3, 5, 7-8), 31-2, 5, t.-p.1.), 51-2, 5, 7-9),

6(1, t.—p.i.), 711-3), 8, 91-2), 10(1),

111-2, 5, 7, t.-p.i.), 24(2), 27, 311-3), 33

Price of this part/Prys van hierdie deel

R5,50

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 44 5

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE

(BOVIDAE, MAMMALIA) FROM SOUTH AFRICA

A. W. GENTRY

British Museum (Natural History), London

(With 28 figures and 4 tables)

[Ms. accepted 10 July 1973]

CONTENTS

PAGE

Introduction ah, Dee Paes mis ‘ ey las

Systematics ; : 4 ; ; 5 2 GS

Comparisons: yen ey Bo NE ae ee Pale

Phenetic comparisons dee ED ceed MSS ML

Conclusion ; : Ser as ; : Bo OIE)

Summary , : ‘ 5 ; 5 Bs iNts5i/

Acknowledgements ‘ : 3 ; IMSS

References Lm May peerage) Mes i eget io

INTRODUCTION

One of the more remarkable mammalian species from the Pliocene deposits

at Langebaanweg in the Cape Province (Hendey 1970a, 19705) is a hitherto

unknown boselaphine. It is only recently that representatives of this tribe have

become known in Africa at all, and this is their most southerly record.

Since reference to the Langebaanweg boselaphine was first made (Gentry

in Hendey 1970a: 114), new material has been recovered and it is now possible

to give a more comprehensive account of its characteristics and affinities. All

the new material is from Bed 2, ‘E’ Quarry, and it is probable that the original

‘E’ Quarry specimens were also from this bed. It is thus an element of a fauna

correlated with 4—-5-million-year-old East African faunas (Hendey 19705, 1972,

1973).

SYSTEMATICS

Family Bovidae Gray

Subfamily Bovinae Gill

Tribe Boselaphini Simpson

Diagnosis

Tendency for skulls to be lower and wider than in other antelopes; horn

cores often with keels but never with transverse ridges, inserted above or just

behind the orbits; cranium little angled on axis of the face; braincase not very

shortened; frontals between the horn bases generally not far above the level

145

Ann. S. Afr. Mus. 65 (5), 1974: 145-188, 28 figs, 4 tables.

146 ANNALS OF THE SOUTH AFRICAN MUSEUM

of the dorsal orbital rims and sometimes with internal sinuses; strong temporal

ridges behind the horn cores; small supraorbital pits; low infraorbital foramen;

median indentation at the back of the palate behind the level of the lateral ones;

posteriorly the ventral edge of the auditory bulla passes upwards to meet the

paraoccipital process; entire occipital surface in one plane facing backwards;

brachyodont cheek teeth; rugose enamel; lower molars generally without

anterior transverse folds (goat folds); long premolar rows; I,s more enlarged

than the other incisors or the incisiform canine; depth of mandible beneath the

molars does not greatly exceed that beneath the premolars. (Modified from

Gentry 1970: 245.)

Remarks

Boselaphini are bush or open woodland antelopes. The only two living

forms are inhabitants of the Indian subcontinent: the nilgai Boselaphus trago-

camelus (Pallas, 1766) and the much smaller four-horned antelope Tetracerus

quadricornis (Blainville, 1816), but there are many fossil species in Eurasia

particularly in the Siwaliks Hills of India and Pakistan, and it is becoming

evident that the tribe has also occurred in Africa in the past. The

Langebaanweg boselaphine will be defined as a new genus and species.

Mesembriportax n. gen.

Type species

Mesembriportax acrae n. sp.

Generic Diagnosis

The type species is the only member of the genus, and a diagnosis is given

under the species.

Mesembriportax acrae Nn. sp.

Holotype

L 13101—a skull with mandibles, and associated vertebral column, rib

fragments and right scapula (Figs 1, 2, 3, 5, 7, 8, 9, 10, 14, 15).

Referred Material

L 20508— Right mandibular fragment with P, to Ms, left mandibular fragment

with M, to Ms, and associated vertebrae, rib fragments, left meta-

carpal fragment, left femur fragment, left lateral malleolus, left

metatarsal, one Ist phalanx and three 3rd phalanges (Figs 10, 11, 21,

22523)

L 20509—Skull fragments including the right horn core (Fig. 4).

L 13106—Skull fragments including parts of the right horn core (Fig. 5).

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 147

L 20506—Skull fragments including left P? and P* to M3, right M! and M2,

left mandibular fragment with M, to Mg, and incompletely erupted

P,, associated with some fragmentary vertebrae.

L 14251—Left and right horn cores (Fig. 4).

S003 4607, b 5692, L 5923, L6587, L 12757, L.12758, L 12812, L 12813,

L 13141, L 13193, L 20234—Horn core fragments.

Upper teeth as follows:

L 6450—Left and right P? and parts of P®; L 10933—left M*; L 10941/6—

right M1 to M® and left P? and M?!; L 12861—right M! and «?; L 12862—left

and right M?; L 13111—right P? and P*; L 13140—right P® and M? and other

fragments; L 14314—left P?; L 14465—left M? and M’; L 20536—left M® and

other fragments.

Mandibular fragments as follows:

L 6601— With left P, to M,; L 11821—with left M, and M,; L 12860—with

right P, to M,; L 13136—with left P,; to M, and part of M,; L 13139—with

right P, to M,; L 14200—with right M, to M, and left P,;; L 14257— with right

P, to M, and left P, and M,; L 20405—with right M, and M, and parts of P,

and M,; L 20534—with right M,, M, and part of M,; L 20538/9—with incom-

plete left P, to M, and right P, to Ms.

Lower teeth as follows:

L 10924—Left M, and M,; L 10936—left P,; L 11979—right P,; L 12698—

right Mg.

The following are juvenile remains:

L 11000/2—Left mandibular fragment with dP, to dP,, M, and M,, and

unerupted P, and P,; right dP,, M, to M;, and unerupted P, to P,.

L 14202/3—Left mandibular fragment with dP, and unerupted M, and left

maxillary fragment with crushed dentition.

L 14237—Left mandibular fragment with dP, and unerupted M,.

L 20688—Left and right mandibular fragments with dP, to dP, and incom-

pletely erupted M,; left maxillary fragment with dP? to dP* and

incompletely erupted M!. This material is possibly associated with

a number of postcranial bones found nearby (Figs 12, 13).

Tentatively referred material

L 13197—Left scapula (Figs 16, 17).

L 13071—Right humerus, radius, ulna, scaphoid, lunate and metacarpal

(Figs 18-22).

L 7625 —Right metatarsal (Figs 21, 22).

L 20334— Right metatarsal.

L 14081—3rd phalanx.

The holotype and all other material is housed in the South African Museum,

Cape Town.

148 ANNALS OF THE SOUTH AFRICAN MUSEUM

Locality

Bed 2 of ‘E’ Quarry, Langebaanweg, Cape Province. In addition, a single

horn core fragment (L 1588A) has been recorded from Baard’s Quarry, Lange-

baanweg.

Age

Pliocene.

Diagnosis

A moderate to large-sized boselaphine; horn cores rather short, inserted

fairly uprightly and widely apart, very divergent basally and slightly less so

distally, strongly compressed medio-laterally and with a postero-lateral keel

and a strong slightly helical anterior keel in their lower part, the anterior keel

being stepped at its top and the succeeding distal part of the horn core being of

small circular cross-section. Frontals extensively hollowed internally, and

their top surface raised much above the level of the top of the orbits; braincase

slightly angled on the face axis; top of braincase not curved downwards pos-

teriorly above the occipital surface; strong temporal ridges on braincase roof

approaching posteriorly and with a rugose surface between them; braincase

widening posteriorly; orbits without a projecting dorsal rim; small supraorbital

pits; nasals long and narrow with large central flanges anteriorly but no lateral

flanges; large preorbital fossa; infraorbital foramen low and situated above the

posterior margin of P?; premaxillae narrow anteriorly but with strong ascending

rami of approximately even width throughout and with a wide contact on the

nasals; palate very wide; median indentation at the back of the palate well

behind the level of the lateral indentations; large mastoid exposure of periotic;

anterior tuberosities of basioccipital fairly wide apart and not very large.

Brachyodont or only moderately hypsodont cheek teeth, with not very

rugose enamel; small basal pillars on upper and lower molars; medial lobes of

upper molars not joined to one another or to the lateral side of the tooth until

late in life; mesostyles quite strong on upper molars; central cavities of upper

molars not very complicated in outline; medial walls of lower molars rather

flat; lower molars sometimes with small goat folds (a transverse flange at the

front of the tooth); long premolar rows with large anterior premolars; paraconid

and metaconid not fused or only just fused on P,; projecting hypoconid on P,;

I,s not greatly enlarged.

Name

The generic name is made from peonpBpivdc (mesembrinos) southern,

and noptagé (portax) a calf, the latter word being a frequent constituent of

boselaphine names. The specific name is from d@Kpa (acra) a cape, and is

given as a Latinized genitive singular.

The Skull

The holotype skull is largely complete, but owing to some distortion in

the facial region it is in three separate parts—the cranium with horn cores, the

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 149

Fig. 1. L 13101, holotype of Mesembriportax acrae. Lateral view of cranium.

Scale in all photographic figures is in centimetres.

palate, and the remainder of the face region. Part of the occipital, the auditory

bullae and other more delicate parts of the basicranium are lost, as is much of

the right horn core and parts of the left. Some pieces, especially of the frontals,

could not be restored to the skull. The dentition is complete except for some -

small pieces missing from the upper and lower cheek teeth.

The horn cores are large, without transverse ridges, not curved backwards,

and inserted above the back of the orbits. They have an ovate cross-section in

their basal part. The anterior keel is very pronounced, and terminates well

below the tip of the core. The torsion in the horn cores is anticlockwise from

the base up on the right side, and it results in the termination of the anterior

keel being situated at what has become the lateral edge of the horn core. Proxi-

mally the keel becomes very sharp-edged, a feature which is best seen in the

specimen L 13106. The postero-lateral edge is also keeled, while the postero-

150 ANNALS OF THE SOUTH AFRICAN MUSEUM

Hl A a a Hu ne il qa

5 2

“

Fig. 3. L 13101, holotype. Dorsal view of nasals with right premaxilla and part of right maxilla;

the arrow points anteriorly.

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 151

medial edge is rounded. There is some individual variation in shape of the

proximal parts of the horn cores; in the holotype and others the medial and

lateral surfaces of the lower part of the horn cores are slightly convex and widely

angled on the posterior surface, which is also convex, but in L 13106 the postero-

lateral keel is more prominent, the posterior surface less rounded, and the

lateral surface flatter. In L 4657 and L 20509 there are well-developed grooves

running along most of the posterior surfaces of the horn cores. Distally from the

termination of the anterior keel the horn core becomes much smaller with a

circular cross-section and in this region the horn cores are less widely divergent.

The relative length of this distal portion appears to vary considerably as can be

seen from Figs 1, 2 and 4.

There are ridges of bone extending antero-medially across the frontals

from the anterior keels of the horn cores. Since this region of the skull is incom-

plete in all specimens it is not known how closely the ridges approach one

another. The supraorbital pits are preserved only on the holotype, where they

are very small. There is almost no sign of a postcornual fossa postero-laterally

to the horn pedicel. The internal hollowing of the frontals extends quite high

into the horn pedicel. Extending posteriorly along the midline of the skull from

the highest point of the frontals between the horn cores is a ridge of bone which

terminates at or near the parieto-frontals suture.

Bohlin (1935) argued that the extinct Eurasian boselaphine Miotragocerus

originally had horn cores of an elongated oval cross-section, and that later

species of the genus evolved periodic growth at the anterior base of the horn core

to produce a series of steps along the course of a sharp anterior keel. The horn

cores acquired greater and greater antero-posterior diameters during the life-

span of an individual. These phenomena may not have occurred in all Miotrago-

cerus, and M. browni and the Samos skulls named M. curvicornis (Andree 1926)

and M. recticornis (Andree 1926) are obviously difficult specimens in which to

visualize them. Thenius (1948) commented that this method of horn core growth

did not necessitate periodic sheddings of the horn sheath provided that sheath

growth was as in other bovids. He thought that a bifurcated sheath like that

of the North American pronghorn, Antilocapra americana, may have occurred

in Miotragocerus. There is no indication of steps along the course of the anterior

Keel in Mesembriportax acrae, except just possibly in L 13106, but one can

scarcely imagine it not having a bifurcated sheath at the point where the anterior

keel disappeared. This, coupled with the very wide divergence of the horn cores,

suggests that males would have fought by grappling, a method of combat

different from the pushing and neck fighting of living Boselaphus (Walther

1958: 358). The four horns of the living Tetracerus quadricornis provide a

precedent for anomalous horn features among the Boselaphini.

The thick frontals with their extensive internal sinuses are one of the most

notable characters of M. acrae. Such sinuses and struts occur in the males of

advanced species of Caprini, and help to protect the brain from the ramming

and clashing of the very large horns (Schaffer & Reed 1972: 47). In M. acrae

(111 DYI UO) GOSOT I PUL (2J2] BY UO) [SZPT “J S919 UIOY JBI JO SMOIA [eIO}EI-OIDIUY “p “Bf

1 nj

hihi tt (tt ttt ttt acta

ANNALS OF THE SOUTH AFRICAN MUSEUM

152

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 153

anterior is

— /otera/

10 mm

L 13106 L 13101

Fig. 5. Cross-sections of two horn cores near their bases.

Fig. 6. Diagram of a bifurcated horn sheath (dashed lines) fitted over a horn core of Mesembri-

portax acrae.

they may have protected the brain against the hypothetical horn prongs but

were more probably connected with the support of the very divergent and over-

hanging horns. Such divergent horns may have needed additional bracing to

prevent them being too easily snapped during the grappling and pushing between

males. The dorso-ventral height of the frontals, irrespective of whether or not

154 ANNALS OF THE SOUTH AFRICAN MUSEUM

they possessed sinuses, may also have been a means of raising the horns above

the level of the orbits, thus giving some protection to the eyes and avoiding

interference with the field of vision.

It is not known whether or not the females of M. acrae had horns; if the

species were like other boselaphines in which the state of this character is known,

they did not.

The temporal ridges of M. acrae arise from the postero-medial edge of the

horn core pedicels, and converge towards the nuchal crest, although they

remain fairly wide apart at their posterior termination. The roof of the brain-

case is flattened in the nuchal region. Arising from the temporal ridges about

midway along their length and extending postero-medially are two swellings

in the bone which converge about half-way along the surface of the braincase.

Anteriorly from these swellings about as far as the parieto-frontals suture, the

roof of the braincase is slightly concave and there is a marked rugosity of the

bone in this region.

The braincase is broad and rather low, and its roof meets the occipital at an

angle of approximately 90°. The nuchal crests are not very prominent; laterally

they are directed slightly anteriorly and terminate in large, somewhat bulbous

paraoccipital processes. The skull is less wide across the occipital surface than

across the orbits. The posterior tuberosities of the basioccipital are fairly

prominent, but the anterior tuberosities are small, and there is a slight narrowing

of the basioccipital anteriorly. A central longitudinal groove is seen only

between the posterior tuberosities, and between the anterior tuberosites there

is even a central ridge. The middle of the basioccipital is not transversely

constricted.

The preorbital fossae are extremely large, and the width of the face on

either side of the narrow nasals is about the same as in the premaxillary region.

There is a marked flaring outwards of the face from beneath the preorbital

fossa to the alveolar margin of the tooth row. Although this is to some extent

a character of all boselaphines, it appears to be exceptionally pronounced in

M. acrae. The palate is very broad, particularly at the molar rows, and the

P*s are in fact stepped in from the level of the M!s.

The upper molars are large and typically boselaphine in morphology,

and all have small basal pillars which are variably developed. In some cases

there are two individual pillars on each tooth and in others only one. In the

holotype the M’s have projections from the most postero-lateral parts of the

cingulum. This feature has not been noted in any other boselaphine specimen,

nor is it present in other Langebaanweg M‘’s. The upper molars are very broad,

the M? being appreciably broader at its base than it is long, the M? being slightly

broader than long, and the M® having its breadth approximately equal to its

length. The joining of the two medial lobes of each upper molar to one another

and to the lateral side of the tooth is possibly more advanced than in earlier

boselaphines; in the holotype the junction is actually made but is very narrow

in each case. The lower molars have small basal pillars decreasing in size from

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 15

yi gr

HII

8 119 2|

Fig. 7. L 13101, holotype. Palate.

M, to Ms. In the latter tooth there is an additional basal pillar between the

middle and the posterior lobes, rather smaller than the main one. The transverse

anterior goat folds are poor on the holotype, but stronger in some other lower

dentitions which are in an earlier stage of wear, e.g. L 11000. There is a large

156 ANNALS OF THE SOUTH AFRICAN MUSEUM

LT puua HE a AAR

Fig. 8. L 13101, holotype. Lateral view of left mandible.

ii i

% Mw

Fig. 9. L 13101, holotype. Occlusal view of left incisors.

metaconid on P, and it approaches closely to the paraconid in the holotype

but without fusion. In L 20508 the fusion has taken place but is not strong. The

central incisors, I,s, are the largest of the symphyseal teeth, but are probably

smaller in relation to I,, I, and C than in the earlier Protragocerus labidotus

from Fort Ternan, Kenya.

The mandibular diastema is about as long as the lower molar row. The

mandibular corpus is relatively shallow and deepens only slightly beneath the

tooth row. The ascending ramus rises steeply, almost at right angles imme-

diately behind the Ms, and its anterior edge and posterior edge below the condyle

are nearly parallel. The two anterior deciduous premolars of both the upper and

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 157

IOUIAUTUTUT UTE ni

HH iu rm tii

CUT AT TAT

ALLA MLL LLL

FULL LL LALLA LLL LLL

Fig. 10. Occlusal view of lower dentitions. Above: L 13101, holotype. Below: L 20508.

Fig. 11. Lateral view of left mandible L 20508.

158 ANNALS OF THE SOUTH AFRICAN MUSEUM

lower dentitions are rather long. The dP? is a little-advanced tooth in which the

shape of the front lobe is quite dissimilar to the back lobe. The dP, has two basal

pillars.

The teeth of M. acrae agree well with other boselaphines and are not very

different from those of Tragelaphini (eland, kudu, bushbuck tribe), so it is

reasonable to suppose that this species fed at least partly by browsing. There is

a great contrast in almost every character with the teeth of grazing antelopes

such as the Alcelaphini (wildebeest and hartebeest tribe).

Measurements in millimetres on the holotype skull are:

Total length of horn core along anterior keel . . . . . . estimated 370

Total length of horn core along medial curve. ; estimated 305

Anterior-posterior diameter at base of horn core at right angles to its longitudinal axis 69,1

Latero-medial diameter at base of horn core at right angles to its longitudinal axis 43,1

Minimum width across lateral sides of horn pedicels 5 . ; ; estimated 145

Length of nasals : ; : estimated 142

Skull width across mastoids behind external auditory meatus ; 5 ; : - ee

Distance between outer edges of occipital condyles . : : . , ; : ad 2!

Width across anterior tuberosities of basioccipital . aa eyes : : : : . ” 268

Width across posterior tuberosities of basioccipital - {- 5° 2 3 eee

Minimum width of palate between medial borders of M's . . . . . =. += 62,5

Occlusal length M'-M?_. ‘ : : ; ; : ‘ ‘ : ; : ; : 76s

Occlusal length M? . 3 3 ; 3 i : : : : : : : : . eee

Occlusal length P?—P* : BEM ee Ce : Shania aie fore MeOneee ss,

Occlusal length P? . d ; ; ‘ 3 ‘ : ; : : : ; ; . RS

Occlusal length P*. ; 4 ; ; 5 : : ‘ ‘ : : : ' : | 0

Occlusal length M,-M, . i : ; : : ‘ é : é : P z » BOF

Occlusal length M, : : : ‘ : 5 ; : : 5 é : : ue ae

Occlusal length M; . : : ‘ : : : : : , : : : : 1 34:6

Occlusal length P.-P, : ‘ ; ‘ : b P ’ ‘ , . : : . = S548

Occlusal length P,_. ; ‘ : ; j : * , 3 : 4 E : Pee! |

Occlusal length P, . 20,4

Overall length of mandible from anterior alveolar margin of ik, to posterior limit of

angle . s 3 3 : : i ‘ 3 : : ; : : estimated 313

Diastema length : : : é : : : : ; estimated 77,0

Depth of mandibular ramus below P, ; : : : : : , ‘ é : J Sao

Depth of mandibular ramus below centre of M, f ; : : P 4 : . 4057

Depth of mandibular ramus below centre of M; ; ‘ : ‘ ; . ) as

Height of ascending ramus between inferior margin and condyle ‘ : : ; > ee

The following additional horn core readings were taken:

Length along Length along Antero-posterior Latero-medial

anterior keel medial curve basal diameter basal diameter

L 4657 — — c.65,0 c.41,0

L 13106 = —— Ole? 40,0

L 14251 c.270 c.230 Tes ml ET

L 20509 c.380 c.310 c.74,0 42,4

The distance between the outer edges of the occipital condyles on L 20509 was

c.74,0.

Measurements on adult upper teeth were:

L 12861 length M? €.2635

L 14465 length M? 27.1

L 20506 length M? 24,9 length P4 13,8

Measurements on lower dentitions are shown in Table 1.

159

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE

== = 0°L6 — = = =S =

S8r'2 =. gE 9°8¢ == 69€ = = =

OSP'2 9'9¢ 7SE = €‘ZE = == ==

C9 aa CE = = = = =

e172 — 0‘0z = = CSI a 8°61

5ST = 7ST = = = = SSI

09959 — ‘es = == = = 6°ES

TSE = ITE ore? — OTE I‘€€ OTE?

CSE 9‘bZ L‘€Z eb LEG 8°SZ Ore’?

v°8L rs L’vl O°€L'9 O'7L9 s‘cu'2 SEL

6ES0¢ TI PESOZ T 80507 I 90507 1 soroz T LSCvl T OOCHI TI GETEL I O98ZI I 6LEITT IZ8IL 1 vc6OI I 1099 T

T A9qeL

Soro — se = =

08€ =e = = =

Lor? = 16 — = Slt

€ZE = = ee OC

74 as ee LE eter

—_— — —-

9JApuod pure ulsreUr JONOJUL

Us9MJOq SNUIeI SUIPUSISE JO JYSIOFY

eA 66 66 66

W (79 66 66

*d MoOjaq Yidep snury

v d 66 66

WW YI8u9] [esn]290

160 ANNALS OF THE SOUTH AFRICAN MUSEUM

Occlusal lengths of deciduous teeth were:

dP2 dP3 dP4 dP2-dP4

L 20688 upper 20,9 21,9 21,4 60,1

L 20688 lower 11,8 18,0 28,1 56,4

L 11000 lower ZT IATL e250 —

L 14202 lower — = Dien —

L 14237 lower — a 25,0 —

Hn ULL ALLL LLL LLL LL nn TULL VLU LILA. LLL LLL LL nn HUAN LL

yr HU ULL UL mint UU AU TUNA LL

7 1

Fig. 12. L 20688, immature left dentitions. Lateral and occlusal views of mandible with dP.—M,,.

Fig. 13. L 20688, immature left dentitions. Occlusal view of maxilla with dP?-dP*.

TUULUUTALV VALS LIL LUELLA LLU LIL

3 114 115 1 rd 118 1

9 2

LLL

0) 211 r]

6 1

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 161

iil

NAH

mn)

HUI I

nil

Fig. 14. L 13101, holotype. Dorsal view of atlas vertebra.

wn HIN

Fig. 15. L 13101, holotype. Lateral view of axis vertebra.

Postcranial skeleton

A iarge number of bovid postcranial bones has been recovered from Bed

2 in ‘E’ Quarry, but only a few were associated with Mesembriportax acrae

cranial remains and thereby positively identified with this species. Apart from

162 ANNALS OF THE SOUTH AFRICAN MUSEUM

Man

(lit

min

Hutt

ual

olf

}

Fig. 16. L 13197, left scapula in lateral view.

See text for explanation of letters in Figs 16-22.

those bones associated with the holotype skull and the cranial elements L 20506

and L 20508, a number of other specimens (L 13197, L 7625, L 20334, L 14081)

can be tentatively referred to M. acrae because they are very similar to positively

identified material. The same applies in the case of a metacarpal which was one

element of an incomplete right limb (L 13071). It is possible that L 13071 and

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 163

iil

Fig. 17. L 13197, left scapula. Glenoid facet in ventral view.

L 13197 belong to the same individual since they were found close together.

Comparison of the postcranial material with extant antelopes is hampered

by the poor samples of the latter— generally not more than two or three indivi-

duals of any one species. On their own, none of the fossil postcranial bones

can be identified at generic or specific level. In many characters they resemble

living tragelaphines, particularly the similarly-sized greater kudu (Tragelaphus

strepsiceros (Pallas, 1766)), and are distinguishable from comparable elements

of alcelaphines. The latter have a morphology suitable for more cursorial forms

in an open plains habitat and Gentry (1970: 277-282) has listed and briefly

discussed limb bone characters to be expected in such cursorial antelopes.

A few cursorial features do appear in M. acrae, and it could have been an -

animal of open woodland. A greater number of cursorial characters appear in

the nilgai, also an animal of open woodland, either because it has a slightly

different ecological niche, or because it has evolved better adaptations in the

last few million years.

The vertebral column associated with the holotype skull is complete as far

as the sacrum, comprising seven cervicals, fourteen thoracics and five lumbars.

In general the preservation of the individual vertebrae is good although many

have lost the processes and other more delicate parts. The atlas and axis are

the least well preserved. The atlas has a well-indented dorsal edge anteriorly

164 ANNALS OF THE SOUTH AFRICAN MUSEUM

216

HUTAULUUHUN

Hill

{

iui

hat

iN HLE

Anil

giz

by?

hall

Hui

w/t il

Fig. 18. L 13071. From the left: lateral view of right humerus, anterior view of same.

for articulation with the skull. It is unlike alcelaphines in having no projecting

point in the middle of the ventral edge anteriorly, and in having convex rather

than concave lateral edges. In these characters it resembles tragelaphines. On

the axis the front edge of the neural spine does not pass very far forwards, and

on the third and fourth cervicals there are widely separated openings of the

vertebrarterial canals. The other vertebrae are not distinctive in any way and

those of the referred specimens (L 20506, L 20508) are essentially similar to those

of the holotype.

The scapulae of the holotype and the referred specimen, L 13197, have quite

small tubera scapulae (a, Fig. 17) in ventral view, with their bases situated above

the level of the rims of the glenoid facets in lateral view (b, Fig. 16). The glenoid

facets have no indentations in their lateral edges (c, Fig. 17). All these characters

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 165

nl il

I Cit

Itt

WAUATANULL

wl vu

PVA MU LAA LLL DA A A

A |

Fig. 19. L 13071, right radius. Left: anterior view. Right above: proximal articular surface.

Right below: distal articular surface.

Arrows point anteriorly.

differ from the condition of alcelaphine antelopes, but the lateral rather than

the central position of the tubera scapulae in ventral view (d, Fig. 17) is not

unlike alcelaphines. The nilgai has a larger tuber scapulae but is otherwise

similar to M. acrae.

The humerus L 13071 has the front of its bicipital groove well anterior

relative to the level of the front of the lateral tuberosity (a, Fig. 18), scarcely

any hollowing for the brachialis insertion under the proximal articular head

(b, Fig. 18), no indentation in the dorsal edge of the distal medial condyle

166 ANNALS OF THE SOUTH AFRICAN MUSEUM

(c, Fig. 18), and no V-shaped ventral projection distally on its lateral side

(d, Fig. 18). It is unlike alcelaphines in all these characters. The medial groove

of the distal condyle is not very deeply incised (e, Fig. 18) and the distal hollow

for the lateral humero-radial ligament is shallow (f, Fig. 18); these characters

are also different from alcelaphines and were considered by Gentry (1970: 281)

to be expected in non-cursorial bovids. However, the wide bicipital groove,

upright distal condyle, and the high distal medial condyle of the fossil are

characters of cursorial bovids. The back of the lateral side of the bone does not

descend low behind the hollow for the humero-radial ligament (g, Fig. 18).

The nilgai agrees with the Langebaanweg fossil only in four characters: the lack

of an indented dorsal edge of the medial condyle, the shallow hollow for the

humero-radial ligament, the wide bicipital groove and the upright distal condyle.

The radius has the back edge of the proximal lateral facet only a little

forward of the level of the back of the medial facet (a, Fig. 19), the proximal

lateral tubercle of small to moderate size and set rather low (b, Fig. 19), the

distal end of the shaft swollen in lateral view, and wide flanges distally on the

anterior surface (c, Fig. 19); all of these characters are unlike alcelaphines and

most of them are unlike cursorial bovids generally. However the distal articular

facets appear quite deeply incised (d, Fig. 19) to receive the proximal row of

carpals, thereby tending to resemble cursorial antelopes. Other characters are a

pointed front edge on the proximal lateral facet (e, Fig. 19), no angled medial or

antero-medial corners on the proximal medial facet (f, Fig. 19), the postero-

medial part of the medial facet projects quite strongly (g, Fig. 19), and the facet

for the cuneiform is wide (h, Fig. 19). The nilgai agrees with the Langebaanweg

fossil in the swollen distal end of the shaft and the wide flanges on the anterior

surface, the deep distal articular facets, the wide cuneiform facet, and the

characters of the outline of the proximal medial facet.

The scaphoid is deep, with an upper edge better indented than in the

greater kudu but less well than in alcelaphines (a, Fig. 20), and it has no posterior

prominence on its medial side (b, Fig. 20). All three characters are unlike

alcelaphines, but the first and perhaps the second agree with the nilgai.

The lunate is without a strongly upstanding projection towards the back

of its upper surface (c, Fig. 20), and has only shallow hollowings on its lateral

(d, Fig. 20) and medial surfaces. Both characters are distinct from alcelaphines

and the nilgai.

The metacarpals of L 13071 and that of L 20508, which lacks the lower half

of the shaft and distal end, have relatively large unciform facets in comparison

with the magnum-trapezoid facets (a, Fig. 22), the total proximal articular

surface does not fill the entire available area at the top of the bone, and at the

top of the distal condyles there are deep tiny hollows on the posterior surface.

These characters are again unlike alcelaphines, and the first and last agree with

the nilgai.

The femur of M. acrae is known only from a small part of the distal end

(L 20508), which is too fragmentary to allow any description of its characters.

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 167

pened

HIN TULL

6 1

FULLER ULLAL CL LLL

ia mn

Fig. 20. L 13071. From above: dorsal view of right scaphoid, lateral view of right scaphoid,

lateral view of right lunate. Anterior sides lie to the right.

168 ANNALS OF THE SOUTH AFRICAN MUSEUM

lil

NULL

mmm

MALATE

with

hi Autinislaninnin

Winey

Hanlin lt

Pea |

D pens

a= = —

ye ies ees

a Cari mr 2 —

Sand SS a

—— T= 3

a pond ——

se —— ===

c= = 2S

a = Ss

s = ———

= —— a

= = —

— = ne —

= Fy

Scere nn — =

a a —

——— — a

———— oe >=

— i es

—— pemeeness ny =

+ = i 2

—=—= ~~ = =

oi. Pecos nan

== CS 4

hs — oS

aoe ee a

= errr —

ny = = n=

TS c= =

= tigen =

= == 7

— = ———

— ee —

— eS —

= = =

a Se =

os = >=

= = =

Fig. 21. Anterior views of metapodials. From the left: L 13071 right metacarpal, L 7625 right

metatarsal, L 20508 left metatarsal.

The innominate and tibia are not represented by any specimens which can be

positively identified with this species and, with the exception of a single lateral

malleolus (L 20508), no tarsal bones are recorded either. The lateral malleolus

is similar to that of the greater kudu in overall size but is narrower and relatively

more elongated antero-posteriorly. It is similar to the lateral malleolus of

alcelaphines but unlike the nilgai in that the posterior end of the tibial facet

curves ventrally and terminates on the posterior surface of the bone.

A secondary, postero-ventrally situated medial facet of articulation with the

astragalus is linked to the main grooved facet but is not visible in the nilgai.

The metatarsals of M. acrae are more like the metatarsals of tragelaphines

than those of alcelaphines. The anterior longitudinal grooves are pronounced

and, although their distal termini are medially situated, they are arched along

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 169

tia mmm ini

Fig. 22. Proximal articular surfaces. From above: L 13071 right metacarpal. L 7625 right

metatarsal. L 20508 left metatarsal.

Anterior edges lie towards the top of the page.

170 ANNALS OF THE SOUTH AFRICAN MUSEUM

HII

a 4

ULLAL LL LL

bes £

LAUT

i mm

iil

ml Mn oT OU lll

Fig. 23. L 20508. 1st phalanx in dorsal view (above) and outer view (below). 3rd phalanx in

dorsal view (above) and outer view (below).

their length towards the lateral edge of the shaft (a, Fig. 21), and their proximal

termini are a little off-centre. The metatarsal III elements are thus a more

prominent part of the anterior surfaces than the metatarsal IV elements. Tragela-

phines, but not the nilgai, also exhibit these characteristics and, although they

might be quite pronounced in the eland, the asymmetry is apparently never as

extreme as in the M. acrae specimens. The secondary (posterior) naviculo-

cuboid facets of the proximal surface are transversely elongated (b, Fig. 22)

as in the tragelaphines and nilgai, and the posterior part of the articular surface

is consequently broader than in alcelaphines and other cursorial bovids (see

Gentry 1970: 280). The posteriorly situated facet for articulation with the vesti-

gial metatarsal is more prominent than usual in tragelaphines but less so than

in alcelaphines, and thus resembles the nilgai. As in the metacarpal, the posterior

hollows above the distal condyles are deep.

The first and third phalanges of M. acrae (L 20508, L 14081) are more

stoutly proportioned than those of the greater kudu but are not as heavily built

as in the eland.

Measurements of the Langebaanweg limb bones are given in Table 2.

LT]

TCE STE 8‘TE ge 9°€€ 6‘0P L‘09 x ' pus [eISIP 18 IOJOUIVIP IOLI9)sod-o19}Ue WINWIXe]

L‘Lb 9°SP ‘Ov mae 70S 69S 719 i : ©: PUd [R}SIP 78 JOJOWURIP osOASURI) LUNWIXEA]

0°97°9 =a SET = 0°87'9 O°€E ‘OE : , : : * -YRYS JO ssouyory} osioAsued) seo]

6‘CP OCP Z8E TE CCE CEE CLL pus [euwxo1d je JojoUWIvIp IOLID\sod-o19;uUK WINWIxeyy

8°6E vip lov 81S 6 8P S‘09 O°SL : : pus [eWIXOId ye JoJoWUIp ss1OAsUR.y CONUUTXB AY

0°€97 O'rS7'2 0°0LZ 0°V97'9 0°997'9 0°067'9 0°097 ; j : ; : F ; : ; Yysug] |/e19AO

a a Se eee

SCOL T VECO? T 80S07 ‘TI [18cl T TLOET T ILO€ET T ILOET T

Tesrejeyoyy = fesieyejoyy«—-esieyeyoy = Fedreovjoyy [ed reovjay snipey sniowinyy

C GPL

AL,

eo)

—Q

aa)

<2)

72 ANNALS OF THE SOUTH AFRICAN MUSEUM

The relative lengths of the radius and metacarpal against the humerus, and

the least transverse thickness across the shaft of the humerus, radius and meta-

podials are shown in Figures 24 and 25. In these figures the Langebaanweg

bones are compared with the following similarly-sized antelopes:

Tragelaphus eurycerus bongo symbol fr

Tragelaphus strepsiceros greater kudu 3 u

Boselaphus tragocamelus nilgai a n

Connochaetes taurinus blue wildebeest -, fe)

Alcelaphus buselaphus hartebeest .. X

The bongo lives in forests, the greater kudu is most typically an inhabitant of

hilly country with thickets preferably near rivers, the nilgai lives in areas with

scrub or open woodland, and the wildebeest and hartebeest are cursorial

antelopes of open plains. In the first three species the humerus is relatively thin

and the radius relatively short, and the bongo has a noticeably short metacarpal

and a thick radius and metapodials. The metapodials of the wildebeest are

thicker than in the hartebeest, and the metacarpal of the hartebeest is notably

long. The proportions of the limb bones thought to be of M. acrae follow fairly

well those of the greater kudu and the nilgai. It is therefore unlikely from this

evidence that the species was either a forest inhabitant or a cursorial plains-

dweller. More probably it lived in an intermediate habitat of thicket, scrub

or open woodland.

COMPARISONS

The tribe Boselaphini belongs to the subfamily Bovinae, which also con-

tains the larger and more specialized Bovini and the African tribe Tragelaphini.

Both the latter probably have a boselaphine ancestry. It is clear that Mesembri-

portax is not a bovine or tragelaphine but a boselaphine which developed some

specializations of its own and retained some different primitive characters from

tragelaphines or bovines. It differs from Tragelaphini by its non-spiralled or

scarcely-spiralled horn cores, their medio-lateral compression, the specialized

distal part of the horn core, the strong temporal ridges, large preorbital fossa,

basioccipital not long or transversely constricted in its centre, and small central

incisors. It differs even from the early or primitive bovines Parabos Arambourg

& Piveteau, 1929 of Europe, Proamphibos Pilgrim, 1939 of the Siwaliks, and

Ugandax Cooke & Coryndon, 1970 of Africa by its smaller size, horn cores with

medio-lateral compression and more upright insertions, distal part of the horn

core with a sharply reduced cross-section, back of the skull less widened, flatter

lateral walls of upper molars and medial walls of lowers, and central cavities

of upper molars somewhat simpler.

It differs from the living nilgai, Boselaphus tragocamelus, by a wealth of

characters, among which it is sufficient to mention the longer horn cores with

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 173

LENGTH u

HUMERUS

300

250

200 LENGTH RADIUS

250 300

LENGTH u

HUMERUS u

300

‘ 4

r = .

& n

250) 2 aie 8

% x

0? 4.

° *

0 wX"x

0 ie hn

200 LENGTH METACARPAL

200 250 300

Fig. 24. Graph of humerus length plotted against radius length and metacarpal length. Scales

are in millimetres. ‘M’ marks Mesembriportax acrae, and other symbols are explained in the text.

ANNALS OF THE SOUTH AFRICAN MUSEUM

174

WSUVIVLIW

0S2

2 x

X 09

nu eo 6°

“Oo C J

00¢

002

0S¢

‘pZ ‘SI Ul se S[OQWAS pUP ‘SOI]JOUI[[IU Ul o1e Sa[eOS

"S9UOg QUIT] INOJ JOJ (SOXE [BOIJADA) S}FEYS OY} JO SOSSOUYSIY} OSIOASULI} JSvO] JSUIeBe (Soxe [e}JUOZIIOY) YISUI] Jo sydeIyH “C7 “BIA

fon Ne ES ee

00% 0S2 002

TWWd¥VOVLAW

n u

n °

00¢ 0S2 002

4y}bua4

SNYaWnH ee

J x X

u x

nn u iy aes 0 (=O 0

” dns C0006 he

n Wo o

0 0

& SSAUMIIY

4sDa7

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 175

an anterior keel ending well below the tip, the horn cores inserted above and

not behind the orbits, the extensive sinuses within the frontals, central cavities

of molars V-shaped rather than curved, flatter lateral walls of upper molars

and medial walls of lowers, longer premolar row and wider premolars, shorter

diastema, deeper horizontal ramus of mandible, and fewer cursorial characters

in the limb bones. Its much larger size and the possession of only two horns

rule out any close relationship to Tetracerus quadricornis. Allometry would

hinder further comparison between these two forms.

Gentry (in Hendey 1970a: 114) had suggested that the first discovered horn

core pieces of Mesembriportax acrae might belong to Miotragocerus or Protra-

gocerus, both of them Upper Miocene boselaphines. The fuller material now

available suggests that comparison of M. acrae should be made with the follow-

ing extinct boselaphines:

Pachyportax Pilgrim, 1937

P. latidens (Lydekker, 1876) of the Dhok Pathan and possibly the

Tatrot Formations, Siwaliks Hills

P. nagrii Pilgrim, 1939 of the Nagri Formation. Siwaliks Hills

Selenoportax Pilgrim, 1937

S. vexillarius Pilgrim, 1937 of the Nagri Formation

Protragocerus Depéret, 1887

P. chantrei Depéret, 1887 from the European Upper Miocene

P. gluten (Pilgrim, 1937) from the Chinji and lowest Nagri Formations

P. labidotus Gentry, 1970 from Fort Ternan, Kenya and dated to

14 million years (Bishop, Miller & Fitch 1969: 685)

Miotragocerus Stromer, 1928

The more familiar name Tragocerus, used until recently for this genus,

was discovered by Kretzoi (1968) to be preoccupied by a beetle, and

Gentry (1971: 284 footnote) suggested Miotragocerus in its place. It

contains the following species and perhaps others:

M. gradiens (Pilgrim, 1937) of the Chinji Formation

M. leskewitschi (Borisiak, 1914) of Sebastopol, Russia

M. amalthea (Roth & Wagner, 1854) from Pikermi, Greece, and ~

elsewhere

M. valenciennesi (Gaudry, 1865) a smaller Pikermi and Samos species

M. browni (Pilgrim, 1937) of the Dhok Pathan Formation

M. spectabilis (Schlosser, 1903) of the Chinese mid-Tertiary

Tragoportax Pilgrim, 1937

T. salmontanus Pilgrim, 1937 of the Dhok Pathan Formation

The comparison of Mesembriportax acrae with the first two genera of this

list is not a matter of difficulty. Pachyportax nagrii is a species of only doubtful

176 ANNALS OF THE SOUTH AFRICAN MUSEUM

validity, having as holotype a hornless female cranium. P. Jatidens is best

represented by a cranium, P./. dhokpathanensis Pilgrim, 1939, in Calcutta, of

which there is a cast in London. From this it can be seen that M. acrae is about

the same size as P. /atidens but has horn cores more strongly compressed medio-

laterally, with a stronger anterior keel, inserted more uprightly and having

greater divergence. The frontals are raised and have more extensive internal

hollowing, the bone surface between the temporal ridges is more rugose, the

orbits are without a dorsal rim, the braincase is more angled on the face axis

and widens posteriorly, and the nuchal crests are weaker.

Selenoportax vexillarius is a large boselaphine among its contemporaries.

The holotype is a cranial roof with horn cores from the Nagri Formation, and

only referred teeth are known from the Dhok Pathan Formation. These teeth,

along with the holotype juvenile maxilla and referred teeth of a supposed second

species S. lydekkeri (Pilgrim, 1910), must be regarded as of uncertain identity.

M. acrae differs from the S. vexillarius holotype by its horn cores being shorter,

more compressed medio-laterally, with the insertion of the anterior keel not

rotated to a medial position, no postero-medial keel, with a reduced circular

cross-section in their distal part, the frontals between the horn cores higher than

the orbits’ dorsal edges, a rugose surface between the temporal ridges, the orbits

without projecting dorsal rims, smaller supraorbital pits, the braincase more

angled on the face axis, and weaker nuchal crests. The state of internal hollowing

in the frontals of S. vexillarius is not known, but the available space must have

allowed much less than in M. acrae. Like Pachyportax, Selenoportax appears to

be an independent boselaphine lineage with no particular relationship to M.

acrae.

The question of the relationship of M. acrae to the last three genera can

only be answered within the context of an assessment of how they are related

among themselves. Protragocerus has been diagnosed by Gentry (1970: 246).

It is an early form with such primitive characters as small size, no internal

sinuses in the frontals, as much as half of the braincase roof curving downwards

posteriorly, no rugosity of the bone surface between the temporal ridges,

prominent nuchal crests, and a small basioccipital with but poor development

of the anterior tuberosities.

P. labidotus is the most completely known species; it has horn cores with a

marked accentuation of their antero-posterior diameter in their lower parts

and hence a strong degree of medio-lateral compression, a clearly demarcated

terminal portion of the horn core above the top of the anterior keel with a much

smaller cross-sectional area, and the horn pedicels becoming extended antero-

medially by the development of ridges. P. labidotus also shows persisting canine

alveoli between the maxillae and premaxillae, and this would probably be seen

also in other species were that part of their skulls known. P. chantrei, the type

species, is very poorly known; it has horn cores with less antero-posterior

elongation and hence less medio-lateral compression and less differentiation of

a terminal portion, and no ridges on the horn pedicels.

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 7

P. gluten, in which Gentry (1970) sank a number of other Siwaliks named

forms, needs further interpretation. The holotype and the conspecific cranium

of Strepsiportax chinjiensis Pilgrim both have horn cores with a less exaggerated

antero-posterior diameter than P. Jabidotus, and are thus less medio-laterally

compressed. The main keel at the back is situated postero-laterally. Gentry

(1970: 257) synonymized Helicoportax praecox and H. tragelaphoides with

P. gluten, but perhaps minimized their differences from the latter. The holotype

partial skull of H. praecox Pilgrim (1937, figs 6, 7, 62), the holotype horn core

of H. tragelaphoides Pilgrim (1939, pl. 4 figs 3, 3a), and two other right horn

core bases, British Museum (Natural History) M. 15469 and M. 15470, all

have larger horn cores than the P. gluten holotype (Pilgrim 1937, figs 12-17).

In their lower parts the strongest keel at the back is the postero-medial and not

the postero-lateral one, and there is more of a transverse ridge across the frontals

between the horn bases. Correlated with the strengths of the two posterior keels,

the medial surfaces of the horn cores are more flattened than the lateral ones.

More distally the postero-lateral keels retain their dominance, and the cross-

sections are closer to that of the P. gluten holotype. The ‘Helicoportax’ specimens

may simply be ontogenetically older than P. gluten, or they may deserve sub-

specific rank as Protragocerus gluten praecox (Pilgrim). Such a subspecies would

not necessarily include the female cranium figured as H. praecox by Pilgrim

(1937, figs 8-11). If the subspecies is valid, it may be a temporal transition from

P. gluten to the Nagri Selenoportax vexillarius as was thought by Pilgrim

himself. It may yet be shown that P. gluten gluten survived to give rise to some

other boselaphine, such as Pachyportax latidens. Sivoreas eremita Pilgrim (1939:

131, pl. 4 figs 1, 1a) may belong to P. gluten. Gentry (1970: 259) supposed it was

antilopine, but was later more doubtful (Gentry 1971: 289). This doubt remains,

but while the holotype frontlet and an assigned horn core piece, British Museum

(Natural History) M. 15495, show a high degree of spiralling, two other horn

core bases, also numbered M. 15495, are rather difficult to distinguish from

Protragocerus. Certainly there is no case for regarding S. eremita as tragelaphine,

which was Pilgrim’s view.

P. gluten differs from P. labidotus by a slightly lower and wider skull,

longer horn cores with no basal enlargement of the antero-posterior axis and

scarcely any differentiation of the distal portion, and no ridges on the pedicels.

Miotragocerus can be told from Protragocerus by its horn cores having less

torsion, being inserted less widely apart, sometimes by its higher and narrower

skulls, and probably (see Thenius 1951: 278) by having hollowed horn core

pedicels. Miotragocerus leskewitschi is a species with a number of primitive

characters. The downwards curvature of the back of the braincase roof (less

pronounced than in Protragocerus), the smallness of the basioccipital and its

poorly developed anterior tuberosities may be mentioned in particular, and the

rather small overall size, short horn cores, and lack of much surface rugosity

between the temporal ridges, may also be primitive. Its nuchal crests are not

strong, and the tip of the horn core has smaller cross-sectional areas than the

178 ANNALS OF THE SOUTH AFRICAN MUSEUM

more proximal part. Whether it is actually early for a Miotragocerus depends on

the dating of Sebastopol, which is discussed in Van Couvering & Miller 1971

and Van Couvering 1972. M. amalthea is a larger species with somewhat longer

horn cores in which the distal portion has a small cross-sectional area, the

frontals have some internal sinuses, and there is an upstanding transverse ridge

between the horn pedicel bases. There is still not a great development of surface

rugosity between the temporal ridges. The basioccipital is larger than in M.

leskewitschi and has scarcely any development of a central longitudinal groove.

The skull is more definitely low and wide than in M. leskewitschi and the horn

cores show some degree of torsion, both of which are approaches to Protrago-

cerus. M. monacensis Stromer and M. pannoniae (Kretzoi) are names which have

been applied to more north-westerly occurring examples of the genus. Illustra-

tions of them (Stromer 1928, fig. 1; Thenius 1948, figs 1, 2; Tobien & Jorg

1959, pl. 11) show short straight horn cores without much torsion but with

terminal portions of abruptly smaller cross-sectional area, stronger rugosity of

the bone surface between the temporal ridges than in M. amalthea, a little

downwards curvatures at the back of the braincase roof, and the premaxilla

rising with even width to make contact with the nasals. M. valenciennesi is

a scarcely known smaller species than M. amalthea, found at Pikermi and Samos,

and there is also perhaps a larger species at these sites (Gentry 1971: 243).

Some Miotragocerus at Samos differ from M. amalthea in having longer horn

cores, with less torsion and most probably with no demarcation of their distal

part; they may also have the braincase more angled on the face axis. They have

received the names M. curvicornis (Andree 1926) and M. recticornis (Andree

1926). A similar skull in the American Museum of Natural History, 20566,

comes from quarry 5 at Samos, which may be later than other sites on that island

(Gentry 1971: 280). M. gradiens is a small, primitive-looking species with some

similarity to M. leskewitschi, but it has a narrower skull and slightly more upright

horn cores, both of which must have helped to distinguish it from the con-

temporaneous Protragocerus. M. vedicus (Pilgrim 1939: 244) appears very

similar to M. gradiens but comes from the later Dhok Pathan Formation;

possibly it is an ontogenetically young specimen of a later species. M. browni

(Pilgrim 1937: 781) is a later Siwaliks species apparently smaller than M. amal-

thea, although Pilgrim (1939: 217) was inclined not to stress this. It may or may

not be conspecific with M. punjabicus Pilgrim, 1910. It has a higher and narrower

skull than the European forms, and the holotype shows long horn cores curving

backwards and without a terminal portion of small diameter. The rugosity of

the frontals is not well marked. M. spectabilis of China is not well differentiated

from M. amalthea.

Tragoportax salmontanus in the opinion of Gentry (1970: 259) includes

material of two other supposed species from the Siwaliks. Tragoportax aiyengari

Pilgrim (1939: 228, fig. 24) was founded on a cranium with horn core bases

said to be from the Dhok Pathan Formation. A plaster cast of this cranium is

available in London, and it differs from 7. salmontanus by being larger, having

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 179

more upright horn core insertions in side view and a less pronounced transverse

raising of the frontals between the horn bases. It does not seem sufficiently

different from T. salmontanus to justify specific rank. Other Dhok Pathan remains

have been referred to T. islami Pilgrim (1939: 230, figs 25d-f, 26). The holotype

is a partial cranium, of which a plaster cast is available in London. The transverse

narrowness of the cranium makes it doubtful that it comes from a Tragoportax

rather than a Miotragocerus, and the remaining horn cores cannot be separated

convincingly from 7. salmontanus. T. salmontanus is smaller than M. amalthea

and has fairly short horn cores probably without a distinct terminal portion,

horn cores inserted close together and with an upstanding transverse ridge of

the frontals between the bases, internal sinuses of the frontals, a strongly rugose

surface between the temporal ridges, the back of the braincase a little down-

curved posteriorly, and a basioccipital with a longitudinal groove extending

forwards to pass between the anterior tuberosities. T. salmontanus is obviously

a member of the Protragocerus-Miotragocerus group, but it is questionable

whether it deserves separate generic rank. It is different from Protragocerus

in the closeness of the horn core insertions, the raised ridge between the pedicels,

the lack of a distinct terminal portion of the horn core, the braincase widening

posteriorly, the rugosity of the bone surface between the temporal ridges, and

the larger basioccipital with its more pronounced central groove. There are

fewer differences from Miotragocerus which is a more diverse genus.

Mesembriportax acrae differs from Protragocerus by its greater size, horn

core insertions less inclined backwards, the greater divergence of the horn cores,

the internal sinuses of the frontals raising them well above the level of the top

of the orbits, no dorsal rim to the orbits, the braincase roof angled and not

curving downwards posteriorly, braincase widening posteriorly, a rugose surface

between the temporal ridges, and a relatively larger basioccipital with stronger

anterior tuberosities. It differs from P. /abidotus by less extreme compression of

the horn cores, the absence of a projecting anterior ridge on the horn pedicels,

temporal ridges wide apart posteriorly, weaker nuchal crests, nasals without

lateral flanges anteriorly, lower molars with straighter medial walls and occa-

sional incipient goat folds, a longer premolar row (Fig. 26), a more massive

metaconid on P,, a closer approach of metaconid and paraconid on P,, and

more reduced canine alveoli at the maxilla/premaxilla junction. Many of these

differences can be seen as evolutionary advances: greater size, frontals’ sinuses,

profile and rugosity of braincase roof, less marked nuchal crests, the basioccipital

characters and the disappearance of the canine alveoli. The tooth characters,

particularly the relative lengthening of the premolar row, cannot be so easily

seen as advances.

Mesembriportax acrae differs from all Miotragocerus by having horn cores

inserted more widely apart and with greater divergence, more extensive frontals’

sinuses, orbits without dorsal rims, and the braincase more angled on the facial

axis. It differs from all except M. amalthea by its greater size, horn cores with

more torsion, and braincase widening posteriorly. It differs from M. amalthea

180 ANNALS OF THE SOUTH AFRICAN MUSEUM

= M

LENGTH Po M4 °

% LENGTH M,-M,

40 50 60 70 80mm

Fig. 26. Graph of length of lower premolar row against lower molar row. a = Miotragocerus

amalthea from Pikermi, e = Protragocerus labidotus, n = Boselaphus tragocamelus, s = Tra-

gelaphus scriptus, u = Tragelaphus strepsiceros, M = Mesembriportax acrae. The upper

diagonal line passes through points along which the length of the premolar row is 70% of

that of the molar row, and the lower one is the corresponding line for 60%.

by having less tendency to a postero-medial keel on its horn cores, and perhaps

more of a rugose surface between its temporal ridges. It differs from M. browni

by the wider skull, shorter horn cores which are less curved backwards, with

less of a tendency to a postero-medial keel, and with the anterior keel terminating

well below the horn core tip. There are other differences of Mesembriportax

acrae from the more primitive-seeming Miotragocerus gradiens and M. leske-

witschi. It is interesting that the plate of a Miotragocerus skull in Tobien &

Jorg (1959, pl. 11) shows a premaxilla rising with even width to make a contact

on the nasals, as in Mesembriportax acrae and Protragocerus labidotus. This

bone conformation may be primitive in Boselaphini.

Definitely identified upper and lower dentitions of Miotragocerus are

known, from which Mesembriportax acrae differs by the straighter medial

walls of its lower molars. The P, of the M. acrae holotype differs in its more

massive metaconid with less differentiation into a neck and strong anterior and

posterior flanges, the paraconid assuming the shape of a low protuberance from

the parastylid rather than a flange, and the hypoconid projecting more strongly

than in many Miotragocerus. However, in L 20508 (Fig. 10) only the

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 181

last difference appears valid. The premolar row is about as long as in M. amalthea

from Pikermi (Fig. 26).

Mesembriportax acrae differs from Tragoportax salmontanus by being

larger, the top of the anterior keel terminating well below the tip on the horn

core, the horn cores inserted more uprightly and further apart, the greater

divergence of the horn cores, the expanded internal sinuses of the frontals and

the orbits without a projecting dorsal rim, the braincase more angled on the face

axis, the braincase roof not curved downwards posteriorly, temporal ridges not

approaching so closely posteriorly, and the central longitudinal groove on the

basioccipital not extending forwards between the anterior tuberosities.

Mesembriportax acrae seems to be a fairly isolated form by the huge extent

of its frontals’ sinuses and by other characters of the skull top and horn cores.

The wide palate may also be notable. It could perhaps have descended from the

Fort Ternan Protragocerus labidotus, which has horn cores with quite strong

torsion, wide insertions, medio-lateral compression, and a long terminal part

of small cross-section. The development of such distinctive horn core tips would

be unique outside Miotragocerus, and only occurred in an area south of the

known range of Miotragocerus. The main problems with such a hypothesis of

descent are what could have happened to the anterior ridges of the horn pedicels

and whether the premolar row could have lengthened. Mesembriportax acrae

seems rather further from Miotragocerus, which has narrower skulls (with the

possible exception of M. amalthea) and more closely inserted horn cores with

generally less torsion. Its wide horn core insertions and the basioccipital mor-

phology suggest that it is not close to Tragoportax salmontanus. A tentative

phylogeny is shown in Figure 27.

Ructiypertex Mesembriportax 5

atidens acrae Tragoportax

salmontanus /

Miotragocerus

Selenoportax ae =

vexillarius

x Protragocerus

gluten

Protragocerus Protragocerus

chantrei labidotus

Eotragus ae

Fig. 27. Phylogeny for some Boselaphini. The time scale is in millions of years. The origins of

Bovini, Tragelaphini and the two living boselaphines are too questionable to be shown here.

ANNALS OF THE SOUTH AFRICAN MUSEUM

182

ae 4 = te “ + + -}- -|- + spremysoeq poAind/s1914sre1s 9SINOD “OH

as = = = = — — — |- + — opdqI][/ e018 SOUDBISAIP “OH

= = = = = = a “te a = oP 9SO[9/9PIM SUOI}IOSUI US9MJOqQ dUISIP “OH

ol str a2 1 ale a ale ar als a = sak/ou S}IqIO pulyeq UONAVSUT “OH

ae = “ = — — -- “f _ + poulpour o10ul/jYsIIdn s10Ul AMAIA OpIS UI UOTIOSUI OH

at i = dt ae Ae zi — — _ quosqe/jusosoid dij Jo UOleoIeWsp ‘OH

i a rn

de als ae a aie ate - + — + + sof/Ou A][eIpoul AIOA Pd}1OSUI [994 IOLIIJUe ‘OH

aT tis = — a 4 + _ = + — Anuepus} sutos/asu0u [904 [eIpou-o019}sod e YIIM “OH

se ae a ale alle oh |. + — — — 97941] /yonu UOISSOIAWIOD [¥19}e[-OIPSlU “OH

ae dle = ake a a = + — + Jasuo]/19}.104s Y1sUZ] 9109 UIOH

aE ate = ale lle 2 + + + + + JoMOIIeU/IOpIM suonsodoid [[nyg

4c _ a ae a = a = fe oo +- IoT[ews /193 Ie] 9ZIS

a a a Sa es Cae a ee es ce me eee a ee a eS

as me aS <5 as aes ae on . © ai a5

Be Be 5 88.62 G8 8s S8 SF S&S a8 ate aie

Se RS Soe Saas Sie as ee SS SSeS

= Ss x7 Se = % a 29 So 6 ZS 6 S56 =x & o>

eee Oe Se Se eS Sse GS Sy Se Se WS

SS 8 ee ew me eee Ss S oS Sse

S = bas = = = = x = = Ay i ®

a > Ss a a >

a) “A % A “” i) G

€ 31981

183

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE

"9109 UIOY = ‘OFy

Or 8b QZ Sb GG aS SP os Or SP SosevjUssI0d 0} po}JOAUOD suOTDRI BUIPII0Ig

Sc SC C7 C7 (6 SZ 6 poreduioo s1aj9e1eY JO ‘ON

BES eae at Fs ee: as we zs ee me ee ee

(++) avon “py wio1y (—) SOOUDIOIP JO “ON

—9

—

—T

™

+ _ | + + + -L 4 if — jussoid/juesqe —_[eyIdId990Iseq uo sA0013 [euIpnyisuoy [esyuag

a. ae a “lL = = = = a -} Joy[euls/ad.1K] [e}Idis50Iseq JO 9ZIg

a + = =I | + — + —- — — 8U01}S/3U01s JOU S}SOIO [PYONN

+ — + “+ _ _ — — + + + UMOP P9AINS/JY4SIVI}s Apo JOOI [eIURID IOIIN}SOg

aE JE = at os = — — — — ou/soA A]IOIIO}SOd suopIM oseoureig

+ — = = — — — — — _ ~ Sso[/9.1OU SIXB 99RJ UO pojsue osvoureIg

ede ck SP at ke ee =“ Jo8.1e]/[]eus syid [eqq1ovsdns Jo ozIg

oS Sr Sa ae eas a as ee ee ee

+ — | - — ~ — “f+ +} -[- ++ 9SO[9/OpIM Sospld [e1oduis) Jo j1ede sourjsiq

+ 4. ae = — =e _— _ — — - JayJOOWIS/ssosn1 SosplI [e1OduIa} usamjoq dovy.Ns ou0g

+ — — — — — — — _ — J3100d/SAIsud} xO SOSNUIS [PUIDJUT YIM s[ej}UOIT

=| = ae ate -- — _ — - — poster 1OU/posiex Soseq “OY U99MjJoq Pasivl s[ejUuOLy

+ + + + + — “+ -f + + 9u0 YVIM/9spiiI JOlI0}Ue Ou Sjoorped “ayy

ae a0 = ap = ar a

2 le _— juasqe 10 9]}11[/JUOSo1d UOISIO} “OH

184 ANNALS OF THE SOUTH AFRICAN MUSEUM

PHENETIC COMPARISONS

A simple quantitative assessment of the phenetic differences of Mesembri-

portax acrae from other boselaphines was undertaken, slightly modified from

the method of Corbet & Hanks (1967: 50). The results of such a straightforward,

unweighted morphological comparison, taking no account of time, phylogeny

or allometry, are interesting in comparison with the findings just’stated above.

Table 3 is a list of non-metrical characters in which the species differ. The

data is taken from material and casts in the British Museum (Natural History)

and from published illustrations. Tooth characters are omitted since they are

available only for Mesembriportax acrae, Protragocerus labidotus, Miotragocerus

amalthea, M. leskewitschi and the nilgai. The state of the characters in Mesem-

briportax acrae has been designated by a + sign, and the opposite state for

each character by a — sign. In the column headed ‘state’, the first alternative

describes the character in M. acrae. No allowance has been made for inter-

mediate states, and blank spaces are left where characters cannot be determined

on available material. Numerical values can then be given to the character

states: + = 1, — = 0, and the sum of differences of M. acrae from each of

the comparative species is shown at the foot of the columns of Table 3, expressed

as a fraction of what the total score would have been if the two species had

differed in every character. Normally this total is 25, but the incompleteness of

some fossils reduces it to 20 or 22. The fractions are standardized as percentages

in the lowest line of the table.

TABLE 4

Percentage differences between pairs of boselaphine species.

1 2 3 4 5 6 u 8 9 40

1 Boselaphus tragocamelus

2 Pachyportax latidens 30

3 Selenoportax vexillarius B74 Iles)

4 Protragocerus gluten 40 35 41

5 Protragocerus labidotus 52° 35 55) 3720

6 Miotragocerus gradiens 59) -56y C79) S32) 2228

7 Miotragocerus leskewitschi 56° 45. 68 24° -20 9

8 Miotragocerus amalthea 44., 35-45) 28 40° 14h ~ 28

9 Miotragocerus browni 40°" 40° 355. 40 60 .)36.° 40) 228

10 Tragoportax salmontanus 48 55° 64 32° 36 27. 24° 28 seee

11 Mesembriportax acrae 48 40 .50 48 52 55 48 28) “4830

The top row of numbers indicates the same species as are listed on the left.

The percentage differences were then found between every pair of species

in Table 3, and the resulting figures are given in Table 4. Figure 28 shows a

dendrogram of phenetic differences based on the percentages in Table 4; in

this dendrogram the positions of the linking lines between any two clusters

represent the mean difference between all members of one cluster and all mem-

bers of the other cluster.

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 185

Several interesting observations arise from these exercises.

1. According to Table 3, M. acrae is phenetically remote from the early

genus Protragocerus, from three of the four Miotragocerus species, from Seleno-

portax vexillarius, and from the nilgai. Percentage differences from all these

forms are 48 or above.

2. It isleast different from Miotragocerus amalthea, being quite pronouncedly

closer to it than to any other species. This is a surprising contrast to the hypo-

thesis of descent from a Protragocerus species. Even if tooth characters were

taken into account, as is possible for these three species, the percentage

differences from M. amalthea and P. labidotus would only change to 31 and 57

respectively. However, Table 4 shows that M. amalthea has no large and no

small percentage differences from any other species. Its total range of readings

extends only from 28 to 45, compared with 9 to 68 for M. leskewitschi or 18 to

79 for Selenoportax vexillarius. It seems to be very much an ‘average’ bosela-

phine. Not only does it have no linkage lower than 28 to any other species, but

it links simultaneously with four species at that level. Its closeness to Mesembri-

portax acrae should be judged only in conjunction with these reservations.

3. Tables 3 and 4 confirm that Mesembriportax acrae is a well-marked

form, and support its attribution to a new genus.

4. Concerning the other species used in the phenetic comparisons and

shown in Figure 28, it may be noted that the species pairs Pachyportax latidens

Boselaphus tragocamelus

5| Pachyportax latidens

Selenoportax vexillarius

Protragocerus gluten

we T labidotus

Miotragocerus gradiens

leskewitschi

T amalthea

rT browni

Tragoportax salmontanus

Mesembriportax acrae

MMe =< ts 0s st a ee ee

50 40 30 20 10 0

Fig. 28. Dendrogram of phenetic differences between some Boselaphini.

186 ANNALS OF THE SOUTH AFRICAN MUSEUM

with Selenoportax vexillarius, the two Protragocerus species, and Miotragocerus

gradiens with M. leskewitschi are convincing or plausible phyletic groupings.

The same may also apply to the further link of the nilgai with Pachyportax

and Selenoportax. This takes place at the 31 level, which is quite a lot lower than

the nilgai’s link with any other species. The complete central cluster joined at

25 is a union of early, primitive or small boselaphines, and this imposes a

‘horizontal’ rather than a ‘vertical’ arrangement of species in the rest of the

dendrogram. It leaves the bottom four species as a cluster of larger and later or

more advanced boselaphines. It is doubtful whether even the two Miotragocerus

species and Tragoportax salmontanus in this later group are phyletically closer

to each other than to M. gradiens or M. leskewitschi in the preceding cluster.

Indeed, 7. salmontanus would have had closer links with M. leskewitschi and

M. gradiens at 24 and 27 respectively had they not already joined Protragocerus.

The inclusion of Mesembriportax acrae in this phenetic group is a clear clash

with the phyletic interpretation presented in this paper, and may be attributed

to parallel evolution and to the fact that Miotragocerus amalthea happens to

link with several species at the 28 level.

The numerical phenetic comparison has been useful in emphasizing the

close morphological resemblance of Mesembriportax acrae to Miotragocerus

amalthea, but the results of such a comparison must evidently be interpreted

very carefully when applied to taxa of different overall body size and different

time levels.

CONCLUSION

Mesembriportax acrae has been seen to have several interesting specializa-

tions, most notably those associated with the frontals and horn cores. It has very

divergent horn cores, possibly with bifurcated sheaths, inserted widely apart

on frontals which possess an extensive system of internal sinuses.

It may have descended from a boselaphine like Protragocerus labidotus,

but in doing so it evolved some cranial characters in parallel with Miotragocerus

amalthea, well known from the Upper Miocene of Europe.

It has been shown that the morphology of the teeth of Mesembriportax

acrae agrees broadly with the boselaphines and tragelaphines, but that the

premolar row is rather long, particularly in comparison with the nilgai. This

might indicate a wholly rather than a partly browsing diet. The limb bone

proportions are nearer to both nilgai and greater kudu than to antelopes of

open plains or thick forests, and the morphology of the limb bones is more like

the greater kudu than the nilgai. So far as can be judged from written comments,

the nilgai may differ ecologically and behaviourally from the greater kudu by

living in terrain which is less hilly and more open, and by being partially a

grazer instead of predominantly a browser (Prater 1965: 273; Schaller 1967: 171;

Dorst & Dandelot 1970: 194; Wilson 1965). It is possible that M. acrae was

nearer to the greater kudu pattern of life. However, as a Pliocene antelope it

could have had less highly evolved adaptations for a nilgai-like ecology.

A NEW GENUS AND SPECIES OF PLIOCENE BOSELAPHINE 187

SUMMARY

Some well-preserved fossils of a new genus and species of a boselaphine

antelope, Mesembriportax acrae, are described from the Pliocene of Langebaan-

weg, Cape Province, South Africa.

ACKNOWLEDGEMENTS

Very extensive help was given to me in the preparation of this paper by

Dr Q. B. Hendey. The photographs were kindly supplied by Héléne Hendey.

I am grateful to Dr Hendey for inviting me to work on such an interesting

antelope as Mesembriportax acrae. | thank the Wenner-Gren Foundation for

Anthropological Research, New York for financial support.

REFERENCES

ANDREE, J. 1926. Neue Cavicornier aus dem Pliocéan von Samos. Palaeontographica 67:

135-175.

ARAMBOURG, C. & PIVETEAU, J. 1929. Note préliminaire sur un ruminant du Pliocéne inférieur

du Roussillon. C.r. somm. Séanc. Soc. géol. Fr. 1929: 144-146.

BisHop, W. W., MILLER, J. A. & Fitcu, F. J. 1969. New potassium-argon age determinations

relevant to the Miocene fossil mammal sequence in East Africa. Am. J. Sci. 267: 669-699.

BOHLIN, B. 1935. Kritische Bemerkungen tiber die Gattung Tragocerus. Nova Acta R. Soc.

Scient. upsal. (4) 9 (10): 1-18.

BorisiAk, A. A. 1914. Mammiféres fossiles de Sebastopol. Trudy geol. Kom. 87: 1-154.

Cooke, H. B. S. & CoRYNDON, S. C. 1970. Pleistocene mammals from the Kaiso Formation

and other related deposits in Uganda. Fossil Vertebr. Afr. 2: 107-224.

CorseET, G. B. & HANKS, J. 1967. A revision of the elephant shrews, family Macroscelididae.

Bull. Br. Mus. nat. Hist. (Zool.) 16: 45-111.

Dorst, J. & DANDELOT, P. 1970. A field guide to the larger mammals of Africa. London: Collins.

GENTRY, A. W. 1970. The Bovidae (Mammalia) of the Fort Ternan fossil fauna. Fossil Vertebr.

Afr. 2: 243-323.

GENTRY, A. W. 1971. The earliest goats and other antelopes from the Samos Hipparion fauna.

Bull. Br. Mus. nat. Hist. (Geol.) 20: 229-296.

HENDEY, Q. B. 1970a. A review of the geology and palaeontology of the Plio/Pleistocene

deposits at Langebaanweg, Cape Province. Ann. S. Afr. Mus. 56: 75-117.

HENDEY, Q. B. 19705. The age of the fossiliferous deposits at Langebaanweg, Cape Province.

Ann. S. Afr. Mus. 56: 119-131.

HENDEY, Q. B. 1972. Further observations on the age of the mammalian fauna from Lange-

baanweg, Cape Province. Palaeoecol. Afr. 6 172-175.

HENDEY, Q. B. 1973. Fossil occurrences at Langebaanweg, Cape Province. Nature, Lond.

244: 13-14.

KRETZOI, M. 1968. New generic names for homonyms. Vertebr. hung. 10: 163-165.

PILGRIM, G. E. 1937. Siwalik antelopes and oxen in the American Museum of Natural History.

Bull, Am. Mus. nat. Hist. 72: 729-874.

PILGRIM, G. E. 1939. The fossil Bovidae of India. Mem. geol. Surv. India Palaeont. indica

(n.s.) 26: 1-356.

PraTeER, S. H. 1965. The book of Indian mammals. 2nd ed. Bombay: Bombay Natural History

Society.

SCHAFFER, W. M. & REED, C. A. 1972. The co-evolution of social behaviour and cranial

morphology in sheep and goats (Bovidae, Caprini). Fieldiana Zool. 61: 1-88.

SCHALLER, G. B. 1967. The deer and the tiger. Chicago: University of Chicago Press.

STROMER, E. 1928. Wirbeltiere im obermiocdnen Flinz Miinchens. Abh. bayer. Akad. Wiss.

32: 1-71.

188 ANNALS OF THE SOUTH AFRICAN MUSEUM

TueNtus, E. 1948. Uber die Entwicklung des Hornzapfens von Miotragocerus. Sber. Gst.

Akad. Wiss. (Abt. I) 157: 203-221.

THENIUS, E. 1951. Ergebnisse palaontologischer Untersuchungen an den Boviden (Mammal.)

des Wiener Beckens. Neues Jb. Geol. Paldont. Mh. 9: 273-282.

ToBIEN, H. & JorG, E. 1959. Die Ausgrabungen an der jungtertiaren Fossilfundstatte Howe-

negg/Hegau 1955-59. Beitr. naturk. Forsch. Siidw Dtl. 18: 175-181.

VAN COUVERING, J. A. 1972. Radiometric calibration of the European Neogene. Jn: BISHOP,

w. Ww. & MILLER, J. A., eds. Calibration of hominoid evolution: 247-271. New York: Wenner-

Gren Foundation; Edinburgh: Scottish Academic Press.

VAN COUVERING, J. A. & MILLER, J. A. 1971. Late Miocene marine and non-marine time

scale in Europe. Nature, Lond. 230: 559-563.

WALTHER, F. 1958. Zur Kampf- und Paarungsverhalten einiger Antilopen. Z. Tierpsychol. 15:

340-380.

WILson, V. J. 1965. Observations on the greater kudu Tragelaphus strepsiceros Pallas from a

tsetse control hunting scheme in northern Rhodesia. E. Afr. Wildl. J. 3: 27-37.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

LLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal article give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BuLLouGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FISCHER, P.-H., DuvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Koun, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 1960b. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L. _

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

A. W. Gentry

A NEW GENUS AND SPECIES OF PLIOCENE

BOSELAPHINE (BOVIDAE, MAMMALIA)

FROM SOUTH AFRICA

VOLUME 65 PART 6 JULY 1974

2.6 6

ANNALS

OF THE SOUTH AFR

ry = 3 ‘3 re ta Pai =) mee v iz = Pte ae

= Gat ah) ae i, a a a

4 f coe

{ a YT Fy

ne Bena rs ae hee de

Let ‘a, 4 ‘1 sa ‘

5 ‘ , ni on - co :

>. a es ‘om xy

7 st . oe fa

q ey

OFAN =) ee KODAK

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 ~~ Band

July 1974 Julie

Part 6 Deel

ORIENTATION AND VARIABILITY IN THE

| OSSICONES OF AFRICAN SIVATHERIINAE

| (MAMMALIA: GIRAFFIDAE)

JOHN M. HARRIS

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 2, 3, 5, 7-8), 31-2, 5, t.—p.i.), 5(1—-2, 5, 7-9),

6(1, t.—p.i.), 701-3), 8, 911-2), 10(1),

11(1-2, 5, 7, t.-p.i.), 24(2), 27, 31(1-3), 33

Price of this part/Prys van hierdie deel

R2,10

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 45 3

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

ORIENTATION AND VARIABILITY IN THE OSSICONES

OF AFRICAN SIVATHERITNAE (MAMMALIA : GIRAFFIDAE)

JOHN M. HARRIS

P.O. Box 40658, Nairobi, Kenya

(With 5 figures)

[MS. accepted 5 July 1973]

CONTENTS

PAGE

Introduction : : : é é , , : ae 189

Posterior ossicones . 3 : : ; : : : 19]

Anterior ossicones . s : : : : ; a SLOT

Discussion . : : : 5 ‘ E 2 : : 197

Summary . : : : , : ‘ , : seu) Gk OT:

Acknowledgements . : ; ; : i ’ b , SEOS

References . : ; , , : ‘ : : : 199

INTRODUCTION

Singer & Boné (1960) placed most of the Plio-Pleistocene sivatheriines from

Africa in the species Sivatherium olduvaiense. In addition they recognized

Libytherium maurusium, founded on a right mandible from Garet Ichkeul, St.

Arnaud (Tunisia) and a second species of Sivatherium—S. cingulatum—based on

dental material from South Africa. There seems to be some controversy over the

correct identification of the common African sivathere, Sivatherium olduvaiense

being used by Hopwood (1934) and Singer & Boné (1960) while Libytherium

olduvaiensis was preferred by Leakey (1965). Arambourg (1960) concluded that

Libytherium maurusium was conspecific with Sivatherium olduvaiense. If Aram-

bourg’s interpretation is correct, the genus Sivatherium (Falconer & Cautley,

1836) has priority over Libytherium (Pomel, 1892) but the species Sivatherium

olduvaiense is a junior synonym of Sivatherium maurusium (Pomel, 1892).

There would, therefore, appear to be two species of Sivatherium from

Africa—S. maurusium and S. cingulatum. A third (and the type) species of

Sivatherium—S. giganteum—is known from the Siwalik Series of India.

Sivatherium giganteum has four ossicones (or horns) in the male, an anterior

conical pair arising from the frontals and a posterior palmate pair situated on

the parietals. The females are believed to lack ossicones. S. giganteum thus

differs from other genera of Asian sivatheres with ossicones. Bramatherium also

has four ossicones—two extending upwards from the fronto-parietal region and

two extending laterally from the parietals but, in contrast to Sivatherium, the

anterior ossicones are much larger than the posterior pair. Hydaspitherium has

one pair of ossicones only, fused at their base into one solid mass, on the fronto-

parietal region. Birgerbohlinia, the only European sivathere, possesses a single

189

Ann. S. Afr. Mus. 65 (6), 1974: 189-198, 5 figs.

190 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. |

A. Skull of Sivatherium giganteum (after Colbert 1935).

B. Skull of Bramatherium (after Colbert 1935).

pair of ossicones that are aligned vertically and are rounded in cross-section.

Because of the incomplete nature of the African sivatheriine material it has

been tempting to orientate the ossicones of the African specimens similarly to

those of the Asian Sivatherium giganteum. The discovery at East Rudolf, Kenya,

ORIENTATION IN THE OSSICIONES OF AFRICAN SIVATHERIINAE 19]

pie

Ree oe

—

\ Sy?

o Se Sap ang a)

Fig. 2

Sivatherium skull from East Rudolf, Kenya. Scale = 10 cm.

of a virtually complete cranium of Sivatherium (Harris in press) has proved of

great importance for reinterpreting incomplete ossicones found at other Plio-

Pleistocene sites in Africa (Arambourg 1948, 1949; Singer & Boné 1960). The

posterior ossicones of the East Rudolf skull are orientated in an entirely different

direction from those of S. giganteum. Re-examination of ossicones from Olduvai

Gorge, Tanzania, has shown that some of the more complete specimens were

orientated in accordance with the East Rudolf skull and it is likely that other

specimens must be similarly aligned. Some doubt is also thrown on whether the

anterior ossicones of the African sivatheriines are in fact separate entities from

the posterior ossicones.

POSTERIOR OSSICONES

Two of the Olduvai Sivatherium ossicones (Old.1.53 of Singer & Boné

(1960) and FLKS 1) extend proximally to include parts of the nuchal crest and

may therefore be confidently orientated in light of the East Rudolf skull. .

Together, these three specimens provide a basis for orientation of other, less

complete ossicones. Four factors appear to be of use in orientation—tuberosities,

grooves, torsion and cross-section shape.

Tuberosities

Although the posterior ossicones of the African sivatheriines appear to be

markedly less palmate than S. giganteum, they are ornamented by a flange or

flanges and a number of discrete knobs. This ornamentation occurs on the

anterior (or distally and, where affected by extreme torsion, lateral) surface of

192 ANNALS OF THE SOUTH AFRICAN MUSEUM

the ossicone. Discrete knobs have been observed diametrically opposite to the

proximal flange in the proximal part of the ossicone but are smaller and fewer in

number than on the anterior surface; distally the knobs appear to be confined to

one edge.

Grooves

Longitudinal or oblique grooves are often present on the ventral surface

of the ossicones. Such grooves differ markedly in development; on some speci-

mens they are very faint, in others they are present on the dorsal surface also.

Where grooves are present on both surfaces the dorsal grooves are always

distinctly fainter and less numerous than those on the ventral surface.

Cross-section

The ventral surface of the ossicone is normally distinctly more convex than

the dorsal surface, especially in the proximal portion.

Torsion

Some sivatheriine ossicones are almost straight, but, where present, torsion

is always clockwise from the base outwards on the right ossicone and anti-

clockwise on the left (cf. Singer & Boné 1960: 494).

Even on incomplete ossicones a combination of the above features normally

serves to orientate the specimen correctly. In view of the above, the following

corrections are necessary to plates illustrating Singer & Boné’s (1960) mono-

graph:

Plate Specimen Side Stated View Corrected View

la Old.2.53 right anteromedial ventral

1b O1d.86 left anteromedial lateral

le Old.1.53 left anteromedial anteroventral

ld Old.2.53 right posterolateral dorsoposterior

le Old.1.53 left posterolateral posterior

hyp Old.86 left posterolateral dorsal

2a Old.3.53 right posterolateral dorsal

2b Old.3.53 right anteromedial ventral

2c Old.3.53 right anterior lateral

3a M.14955 right posterodorsal dorsal

3b M.14955 right anteromedial ventral

3c Old.52 SHK/BKII+ M.145945 right anteromedial ventral

3d Old.52 SHK/BKII+ M.145945 right posterolateral dorsal

29c C.431A right anterior anterior

29d C.431A right anteromedial ventral

43a Hopefield 4372 right posterolateral dorsal

43b Hopefield 4372 right anterior anterior

43c Hopefield 4372 right anteromedial ventral

44a Hopefield 4373 left anteromedial ventral

44b Hopefield 4373 left anterior anterior

44c Hopefield 4373 left posterolateral dorsal

Sla St Arnaud, Algeria 1948-1-2 right anteromedial anteroventral

516 St Arnaud, Algeria 1948-1-1 right medial dorsal

ORIENTATION IN THE OSSICONES OF AFRICAN SIVATHERIINAE 193

Once correct orientation has been achieved, three major shapes are apparent

in the posterior ossicones of the specimens from Olduvai and appear to be

typical of ossicones from other African sites also.

Type A

The posterior ossicone extends outwards and backwards, then perhaps

slightly upwards. Flanges and knobs are present on the anterior and lateral

surfaces. Torsion is only slight (Fig. 3).

Olduvai examples: FLK S 1, Old.63 BK II 431, Old.5.53, M.14955, M.149545.

Type B

The ossicone extends outwards, then upwards, then inwards and forwards.

The distal tip of the ossicone is thus bent forward and sited in front of the

anterior flanges on the proximal portion of the ossicone. Flanges and knobs are

sited on the anterior (and, through torsion, outer) edge of the ossicone, but a

knob is also present on the posteroventral edge (diametrically opposite to the

proximal flange) in the proximal portion of the ossicone (Fig. 4).

Olduvai examples: BK II 068/5746 (= Old.1.53), BK II 068/5747 (= Old.

2.53), BK II 068/5753.

Type C

Recognizable parts of the cranium have not been observed attached to

ossicones of this type. These ossicones are straight and, it is believed, point

directly backwards. Knobs may be present on both medial and lateral edges but

are more abundant (and more prominent) on the lateral edges (Fig. 5).

Olduvai examples: Old.68 BK II 8.34, Old.53 BK II 86.

Singer & Boné (1960: 494) commented on the variability of the posterior

ossicones from Olduvai and suggested that the ossicones exhibiting extreme

torsion might possibly represent male specimens. Colbert (1935: 342) agreed

with earlier workers that Sivatherium giganteum lacked ossicones in the female.

It is likely that this also applied to other species of Sivatherium and, if this view

is correct, all African sivatheriine ossicones are from male animals. Disparity in

size and development of the knobs and flanges may represent individual varia-

tion, but it is also possible that the amount of secondary bone apposition, in the

form of ornamentation, might be a function of the age of the individual.

All the Sivatherium posterior ossicones so far collected from Olduvai fall.

into one of the three morphological groups listed above. It may be that these

three groups possess taxonomic significance and certainly the two pairs of

associated left and right ossicones (Old.1.53+Old.2.53 and FLK S 1) are sym-

metrical. Of the two posterior ossicones on the East Rudolf skull, however, the

tight ossicone falls into Group A and the left into Group B. Material collected

from Olduvai subsequent to the publication of Singer & Boné’s monograph

includes a large number of sivatheriine postcranial elements. These are currently

being investigated and it will be interesting to see whether the presence of more

than one species of Sivatherium is indicated.

194 ANNALS OF THE SOUTH AFRICAN MUSEUM

ress

Fig. 3

Restoration of skull of Sivatherium maurusium Type A.

A. Dorsal view. B. Lateral view. Scale = 10 cm.

ORIENTATION IN THE OSSICONES OF AFRICAN SIVATHERIINAE 195

——_—_———=

Quy oe

B ecules

Fig. 4

Restoration of skull of Sivatherium maurusium Type B.

A. Dorsal view. B. Lateral view. Scale = 10 cm.

196 ANNALS OF THE SOUTH AFRICAN MUSEUM

Wu __F

Fig. 5

Restoration of skull of Sivatherium maurusium Type C.

A. Dorsal view. B. Lateral view. Scale = 10 cm.

ORIENTATION IN THE OSSICONES OF AFRICAN SIVATHERIINAE 197

ANTERIOR OSSICONES

Singer & Boné (1960: 492) described anterior ossicones of Sivatherium

maurusium from Hopefield and Tierfontein. Anterior ossicones are present on

the East Rudolf skull and on two Olduvai specimens (Old.68 BK II S.34 and

FLK § 1). Anterior ossicones in Sivatherium giganteum are discrete forward-

projecting conical structures sited above the orbits. Those of the East African

specimens appear to be flange-like ossifications sited on the lateral edges of the

cranial vaults behind the orbits. There is some doubt whether the ‘anterior

ossicones’ of the East African sivatheriines can be interpreted as discrete

structures or merely represent an anterior extension, or the beginning, of the

posterior ossicones. None of the East African ‘anterior ossicones’ are sculpted on

one side by the grooves that are present in the Tierfontein specimen (C.431B)

and, as such grooves are typical of the posterior ossicones, there must be some

doubt about the correct identification of the Hopefield fragment.

DISCUSSION

Quite apart from the distinct possibility of further taxonomic subdivision

on the basis of ossicone morphology and postcranial elements, some doubt must

also be placed on the current identification to generic level of the African Plio-

Pleistocene sivatheriines. As is recorded elsewhere (Harris in press), the skull of

Sivatherium giganteum has a much shorter facial region and a deeper and wider

cranial region than the East Rudolf skull. The posterior ossicones of the African

sivatheriines are orientated differently and the anterior ossicones, if interpreted

as such, are sited more posteriorly and are less prominently developed. The teeth

of the African and Asian sivatheriines are, however, similar in morphology.

It is possible that Sivatherium maurusium and S. cingulatum may be only

specifically distinct from S. giganteum. It is also possible that, because the Asian

sivatheriine genera overlap considerably in morphology except for their ossi-

cones (Singer & Boné 1960: 520), the different skull shape and ossicone orienta-

tion of the African specimens may warrant generic distinction from their Asian

relatives. Such major taxonomic realignment must, however, await the results of

current investigations into the postcranial anatomy of the African sivatheriines,

revision of the Asian forms, and, if possible, retrieval of further specimens from

African sites.

SUMMARY

Isolated ossicones of Sivatherium maurusium can be orientated by means of

their ornamentation, torsion and cross-section. Three basic shapes of posterior

ossicone have been observed. The presence of a well-developed anterior ossicone

in this taxon is doubtful.

198 ANNALS OF THE SOUTH AFRICAN MUSEUM

ACKNOWLEDGEMENTS

I am most grateful to Professor C. S. Churcher for information concerning

the taxonomy of African sivatheres. The manuscript was kindly read and

criticized by Dr M. G. Leakey and Professor Churcher.

REFERENCES

ARAMBOURG, C. 1948. Un sivatheriine nord-africain: Libytherium maurusium Pomel. C.r.Séance-

Soc. géol. Fr. 1948 (séance du 10 Mai): 178-179.

ARAMBOURG, C. 1949. Les gisements des vertébrés villafranchiens de l’Afrique du Nord. Bull.

Soc. géol. Fr. (5) 19: 195-203.

ARAMBOURG, C. 1960. Precisions nouvelles sur Libytherium maurusium Pomel, giraffide du

Villafranchien d’Afrique. Bull. Soc. géol. Fr. (7) 2: 889-894.

CoLserT, E. H. 1935. Siwalik mammals in the American Museum of Natural History. Trans.

Am. phil. Soc. (n.s.) 26: i-x, 1-401.

Harris, J. M. (dn press.) Pleistocene Giraffidae (Mammalia, Artiodactyla) from East Rudolf,

Kenya. Fossil Vertebrates Afr. 4.

Hopwoop, A. T. 1934. New fossil mammals from Olduvai, Tanganyika Territory. Ann. Mag.

nat. Hist. (10) 14: 546-550.

FALCONER, H. & CAUTLEY, P. T. 1836. Sivatherium giganteum, a new fossil ruminant genus from

the valley of the Markanda, in the Siwalik branch of the sub-Himalayan Mountains.

Asiat. Reschs 19: 1-24.

LEAKEY, L. S. B. 1965. Olduvai Gorge 1951-1961. 1. A preliminary report on the geology and

fauna. Cambridge: University Press.

PomeL, A. 1892. Sur le Libytherium maurusium, grand ruminant du terrain pliocéne plaisancien

d’Algérie. C.r. hebd. Séanc. Acad. Sci., Paris 115: 100-102.

SINGER, R. & Bong, E. L. 1960. Modern giraffes and the fossil giraffids of Africa. Ann. S. Afr.

Mus. 45: 375-548.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FIscHER, P.-H., DUVAL, M. & RarFFy, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 1960b. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

John M. Harris

ORIENTATION AND VARIABILITY IN THE

OSSICONES OF AFRICAN SIVATHERIINAE

(MAMMALIA: GIRAFFIDAE)

VOLUME 65 PART 7 JULY 1974

Be OweGe

ANNALS

"OF THE SOUTH AFRICAN —

CAPE TOWN

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 Band

July 1974 Julie

Part 7. Deel

meee STATUS OF THE PLIO-PLEISTOCENE PANOPEA

IN SOUTHERN AFRICA

(MOLLUSCA, BIVALVIA, HIATELLIDAE)

BRIAN KENSLEY

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 201, 3, 5, 7-8), 3(1-2, 5, t.-p.i.), 51-2, 5, 7-9),

6(1, t.—p.i.), 711-3), 8, 9(1-2), 10(1),

11(1-2, 5, 7, t.—p.i.), 24(2), 27, 3101-3), 33

Price of this part/Prys van hierdie deel

R3,30

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 46 1

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN

AFRICA (MOLLUSCA, BIVALVIA, HIATELLIDAE)

BRIAN KENSLEY

South African Museum, Cape Town

(With 8 figures and 1 map)

[Ms. accepted 7 August 1973]

CONTENTS

PAGE

Introduction : 3 ? : rf ; : whe LOO

Systematic discussion ; ; ; ‘ : i 200

Material and measurements . : ; ; oo ZO.

Southern African localities . : ; : - 206

Mediterranean and West African localities . AUS)

Description . ‘ : : : : é : . 209

Discussion . s : , : : i : ee 20

iBnietecologicaliobsenvations’ = 2 94)... 2. 210

Summary . : ; é ‘ 3 ; ‘ sf e202

Acknowledgements) om) ei) ee eae ey a DD

References . E ; ; 5 : : ; 213

INTRODUCTION

The bivalve genus Panopea is represented by about 10 living species in

temperate to subtropical seas. All are inhabitants of sand or mud, being buried

to depths of 1-2 metres below the substrate surface, in water depths varying from

the Low Water of Springs level, to deeper water. The live animal communicates

with the overlying water by means of elongate siphons. The North American

representative, P. generosa Gould, occurs in large numbers in the region of

Puget Sound, where, under the name ‘geoduck’, the living animals are eagerly

sought after as culinary delicacies (Phillips 1970). Elsewhere, other species are

apparently not as common.

As a fossil form, the genus is known from Lower Cretaceous to Recent ©

times, and is represented by numerous species. In South Africa, P. gurgitis

(Brongniart) is known from the Lower Cretaceous (Neocomian) of Uitenhage

and Zululand.

Regarding younger forms from South Africa, Stow (1871), reporting on a

raised beach in the Port Elizabeth district, mentioned the presence of a fossil

Panopea. Earlier, in 1855, Woodward proposed the name P. natalensis for a

specimen erroneously recorded from Natal but in fact coming from Angola.

Since then several more specimens have been recorded from various sites, under

a variety of names. Woodward (1855) mentioned the similarity of P. natalensis to

199

Ann. S. Afr. Mus. 65 (7), 1974: 199-215, 8 figs, 1 map

200 ANNALS OF THE SOUTH AFRICAN MUSEUM

the Mediterranean P. aldrovandi, while Barnard (1964: 559), using the name

P. aldrovandi for Pleistocene forms, raised the following question: ‘Are there any

clear-cut characters apart from minor and inconstant differences in shape due to

growth changes, which will serve to differentiate these so-called species ?’

Since more Pleistocene specimens have become available, it was thought

necessary to establish the systematic position of this species beyond any reason-

able doubt. With this in view, the present investigation was carried out. Speci-

mens, both fossil and recent, were obtained from as many sources as possible

(Table 1).

It must be noted that the exact age of many of the South African ‘beach’

deposits, referred to either as early Pleistocene or late Tertiary, is still uncertainly

determined, and that the term ‘Plio-Pleistocene’ is deliberately used in the present

work in its broadest sense.

SYSTEMATIC DISCUSSION

Phylum MOLLUSCA

Class BIVALVIA

Order MYOIDA

Family Hiatellidae

Genus Panopea Ménard de la Groye, 1807

Apart from the characteristics of the family, the genus Panopea is charac-

terized by the possession of a large ligamental nymph, a wide pallial sinus, and

valves gaping at both ends.

Subgenus Panopea

The subgenus Panopea is characterized by a continuous pallial line, and a

single cardinal tooth in each valve.

Panopea (Panopea) glycymeris (Born)

SYNONYMY

Mya glycymeris was the name first used in a valid description of the present

species and is given first in the synonymy.

Panopea aldrovandi Ménard, 1807, was the first usage of the valid generic

name for the present species (Mya and Glycimeris both having been pre-

occupied), and is given next in the synonymy. This is followed by the correct

designation, i.e. Panopea glycymeris. As a proliferation of names has been used

with reference to the present species, the rest of the synonymy has been given

alphabetically, with entries under each name listed chronologically.

Mya glycymeris Born, 1778: 10; 1780: 20. Chemnitz, 1782: 33. Gmelin, 1790:

3222. Montagu, 1808: 19. Turton, 1819: 107. Wood, 1825: 13.

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 201

Fig. 1. Panopea glycymeris: External view.

a. Sicily; b. Portugal; c. Mocamedes; d. Baia dos Tigres; e. Velddrif; f. Klein Brak River.

Panopea aldrovandi, Ménard, 1807: 136; 1808: 464. Nicklés, 1950: 228. Franca

1960: 37. Barnard, 1964: 557.

Panopea glycymeris: Lamy, 1925: 267. Thiele, 1935: 924. Imperatori, 1961: 141.

Ghisotti & Steinmann, 1969: Sheet 78—Ac-—O1. Treatise, 1969: N700.

Glycimeris aldrovandi: Pallary, 1900: 410; 1920: 94.

Glycimeris glycimeris: Gray, 1847: 189. Tryon, 1869: 60. Malatesta & Nicosia,

19552 177.

Glycimeris rugosa: Bosc, 1802: 5. Adams & Adams, 1856: 350.

Glycymeris aldrovandi: Fischer, 1880-1887: 1125.

202 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 2. Panopea glycymeris: Internal view.

a. Sicily; b. Portugal; c. Mocamedes; d. Baia dos Tigres; e. Velddrif; f. Klein Brak River.

Panopae aldrovandi: Lamarck, 1818: 457.

Panopaea aldrovandi: Blainville, 1825-27: 571. Deshayes, 1832: 698; 1835: 67.

Philippi, 1836-44: 6, 7. Valenciennes, 1839: pl. 1. Deshayes, 1843: 133.

Potiez & Michaud, 1844: 259. Forbes & Hanley, 1853: 178. Wood-

ward, 1855: 218. Sowerby, 1873: pl. 1. Debeaux, 1884: 510. Clessin,

1895: 46. Dautzenberg, 1910: 148; 1912: 99.

Panopaea attenuata Sowerby, 1893: pl. 3; 1889: 156. Clessin, 1895: 45.

Panopaea australis Valenciennes, (non Sowerby), 1839: 3, 34; 1843: pl. 8, pl. 12.

Woodward (non Sowerby), 1855: 220.

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 203

Panopaea cyclopana Di Monterosato, 1889: 26. Locard, 1892: 256.

Panopaea faujasii: Philippi, 1836-44: 6, 7. Valenciennes, 1843: pls. 2, 3, 4, 6.

Wood, 1853: 283. Chenu, 1862: 27. Sowerby, 1873: pl 2. Clessin,

1895: 51.

Panopaea glycimeria: Locard, 1886: 384.

Panopaea glycimeris: Brown, 1827: pl. 10. Deshayes, 1843: 5. Brown, 1844: 112.

Petit de la Saussaye, 1869: 33. Mayer, 1870: 24, 39. Di Monterosato,

1875: 19; 1878: 75. Crosse, 1884: 191. Petersen, 1888: 159. Monchar-

mont Zei, 1960: 184. Di Monterosato, 1889: 26. Nobre, 1938-40: 712.

Ruggieri, 1943-44: 99; 1948: 70; 1949: 92.

Panopaea glycymeris: Turton, 1822: 42. Brown, 1827: pl. 10. De Gregorio, 1929:

5. Ekman, 1953: 84. Priolo, 1966: 669.

Panopaea glycymeris faujasi: Chiesa, 1932: 171.

Panopaea natalensis: Woodward, 1855: 220. Smith, 1906: 294. Schwarz, 1910:

115. Barnard, 1951: 167.

Panopea dreyeri Van Hoepen, 1940: 186.

Panopea glycimeris: Winckworth & Winckworth, 1935: 162. De Stefani, 1941:

274.

Panopea glycymeris var. rugosa Lamy, 1925: 269.

Panopea natalensis Woodward, 1855: 220. Smith, 1903: 399. Lamy, 1925: 271.

Barnard, 1962: 187.

Panope faujasii Ménard, 1807: 115.

MATERIAL AND MEASUREMENTS

A complete list of the material examined is given in Table 1. Only shell

characters were used for comparative purposes. These characters include hinge

structure, muscle and pallial scars, and general proportions. For the comparison

of general proportions, the following measurements were used:

greatest diagonal length (a)

umbo to ventral margin (b)

umbo to antero-dorsal corner (c)

umbo to postero-ventral corner (d)

Figure 3 illustrates these dimensions, which were taken for all measurable

specimens. In the case where both valves of the shell are preserved, only one

valve was measured. Figure 5 illustrates the scatter resulting from the plotting

of diagonal length against the distance from the umbo to the ventral margin.

Figure 6 illustrates the scatter obtained from a comparison of the distance from

the umbo to the antero-dorsal corner with the distance from the umbo to the

postero-ventral corner. (Because of the very gradual curve of the postero-dorsal

margin, umbo to postero-dorsal corner is difficult to measure, and was not used.)

204 ANNALS OF THE SOUTH AFRICAN MUSEUM

Table 1

Umbo to Umbo to

Greatest Umboto antero- postero-

diagonal __ ventral dorsal ventral

length margin corner corner

Specimen Locality Valves mm mm mm mm

Paris Museum Mediterranean L&R 247 127 105 170

Paris Museum Sicily L&R Die. 146 114 189

Paris Museum Portugal L 211 120 104 131

British Museum Malta L&R 246 132 106 170

S.A. Museum Mediterranean R 173 88 71 115

Lisbon Baia dos Tigres L 204 126 96 121

Lisbon Mocamedes R 167 92 1 112

Geological Survey Velddrif R 240 132 112 154

S.A. Museum

K4391 Velddrif L&R 234 133 95 159

K4390 Velddrif L&R 244 133 Li 158

K4393 Velddrif L&R 202 109 87 137

K4394 Velddrif L&R 198 114 88 129

K4395 Velddrif L&R 228 131 100 155

A2139 Velddrif L&R 209 111 82 150

Geological Survey Klein Brak River L&R 185 95 84 121

National Museum

D2010a Klein Brak River R 205 116 83 143

D2010b Klein Brak River R 194 103 84 128

D2010c Klein Brak River IL, 178 95 76) 117

D2010d Klein Brak River R 183 99 74 125

D2010e Klein Brak River L 155 84 61 104

D2010f Klein Brak River L 162 92 71 113

D2010g Klein Brak River R 209 104 90 133

S.A. Museum

K4385 Klein Brak River L&R 183 99 Ti 122

K4389 Klein Brak River L&R 165 93 mn 112

K4384 Klein Brak River L&R 180 98 71 125

K4383 Klein Brak River L 197 118 83 134

K4386 Klein Brak River L&R 168 83 65 115

K4388 Klein Brak River L 184 103 76 132

3486 Klein Brak River L&R 157 91 68 119

The following material, being damaged, could not be measured:

S.A. Museum. Velddrif L & R, L, R

Geological Survey. Velddrif 2 fragments

Paris Museum. Dakar L

The possibility was considered that shell porportions changed differently with

growth in the different populations. With this in mind, 2 individuals were

selected from each of the Mediterranean, Velddrif and Klein Brak populations,

and 6 growth lines traced on each of these. The length/height ratios were plotted

for each of these growth lines and are shown in Figure 7. These dimensions are

given in Table 2. In each case, the final length/height set represents the total

dimensions of the specimen. For convenience, height was measured, not from the

umbo, but from the most dorsal point of the shell as illustrated in Figure 4.

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 205

Fig. 3. Diagram to illustrate dimensions used. Fig. 4. Diagram to illustrate

method of growth-line measure-

ment.

150

100

Breadth mm

Mediterranean

Angola

Velddrif

Klein Brak R

300

ANNALS OF THE SOUTH AFRICAN MUSEUM

io)

Umbo to antero-dorsal angle

50 100 150 200 250

Umbo to postero-ventral angle mm

Fig. 6

Table 2

Specimen Locality Growth-line Measurements: Length x Height mm

Paris Museum Sicily 48 x 31 102x 65 165 x 100 197 x 118 236 x 145 272 x 159

British Museum Malta 24 x 16 78 x 46 113x 75 177 x 109 200 x 122 246 x 144

Geological Survey Klein Brak River 41 x 25 75x 46 109x 64 141 x 84 170x 93 185 x 103

S.A. Museum

K4385 Klein Brak River 27 x 16 59 x 38 95x 58 147x 88 167 x 98 183 x 102

K4391 Velddrif 52x 31 109 x 66 147x 87 182 x 109 210 x 128 234 x 145

K4390 Velddrif 56 x 34 88 x 58 121x 76 189 x 114 213 x 135 244 x 145

SOUTHERN AFRICAN LOCALITIES (See map)

Zwartkops, Port Elizabeth

Originally designated as Pliocene or Post-Pliocene (Stow 1871) it has been

suggested that these formations are of Pleistocene age (Schwarz 1910).

Klein Brak River, Mossel Bay (Rogers 1906; Schwarz 1910)

These deposits have a similar age to the previous locality, i.e. Plio-Pleisto-

cene.

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 207

300

Klein Brak R

250

Mediterranean

Velddrif

Og oe <1 <

150 200

Length mm

100

WW Y}peag

Fig. 7

200

150

208 ANNALS OF THE SOUTH AFRICAN MUSEUM

Morocco

Dakar

Mocamedes

Baia dos Tigres

Port Elizabeth

MAP 1. Locality map for Panopea glycymeris.

Velddrif

Situated near the mouth of the Great Berg River, the farm ‘Kruispad’, 8 km

from the present seashore on the west coast, has been given a Pleistocene age

(Visser & Schoch 1973).

Meob Bay, South West Africa

The exact origin of these specimens is uncertain, as they were washed ashore

at springtide.

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 209

Baia dos Tigres, Angola (Franca 1960)

Whether this specimen is from a recently living animal, or a fossil from a

raised beach deposit, is difficult to establish. In appearance, state of preservation,

it is not unlike some of the material from Klein Brak River.

Mocdmedes, Angola (Franca 1960)

This is a valve from a recently living animal, as a portion of the external

ligament is preserved.

MEDITERRANEAN AND WEST AFRICAN LOCALITIES

Atlantic coasts of Portugal and Spain; Mediterranean coasts of Spain,

France, Italy, Morocco; Balearic Islands; Sicily; Malta; Port of Dakar, Senegal.

DESCRIPTION

Shell equivalve, elongate, roughly trapezoidal, gaping at both ends, strongly

vaulted. Umbo situated slightly anterior to midpoint, just below dorsal vault.

Hinge axis and dorsal margin more or less straight. Anterior margin slightly

concave, sloping posteriorly in the ventral region. Antero-dorsal corner evenly

rounded. Posterior margin truncate, postero-dorsal and postero-ventral corners

smoothly rounded. Hinge bearing slight conical protruding cardinal tooth in

each valve, situated just below umbo. Tooth of right valve fitting into hollow

anterior to cardinal tooth of left valve. Cardinal tooth bearing a fine mid-dorsal

ridge. Posterior to tooth, a stout, solid, elongate roughly rectangular structure.

Nymph narrow near umbo, widening posteriorly.

Interior of valves with roughly oval anterior adductor muscle scar, posterior

broad, irregular, but always continuous. Pallial sinus variable, triangular to

rounded. One to several thin lamellar structures sometimes developed in dorsal

half of valve, usually running obliquely from internal dorsal margin, alongside

adductor muscle scar, more frequently seen in shells from Klein Brak River,

function possibly to give added strength to strongly vaulted shell. Margin of

shell often eroded, revealing layered structure; occasionally, perhaps due to

injury, more extensive layers of calcareous material becomes separated from rest

of surface of shell.

Exterior of valves bearing rounded concentric ridges, smoother and more

regular at early growth stages. Ridges and growth lines becoming irregular with

age. Overall external surface smoothness varying with individuals. Fine short

oblique striae sometimes developed in region of ventral margin.

While the material from the Velddrif area is usually bleached white and

often fractured, the material from Klein Brak River usually has a remarkably

fresh appearance, and could be mistaken for living, were it not that the localities

in the raised beaches are so well recorded. In both these Incalities, specimens with

both valves preserved in the ‘living’ position are sometimes found.

210 ANNALS OF THE SOUTH AFRICAN MUSEUM

DISCUSSION

Comparison of the hinge lines of specimens from the different localities

failed to reveal any differences either in size or disposition of the structures

(Fig. 8). (Note: the protruding cardinal tooth tends to be broken off.)

The pallial line and pallial sinus both appeared to be variable between

individuals from the same locality and therefore of no diagnostic value. From

Figures 5 and 6, it would be difficult to separate any populations; indeed, both

figures would seem to indicate a single continuous ‘population’. The lines

obtained from growth-line measurements (Fig. 7) would seem to indicate that

the pattern of growth is similar in the Klein Brak River, Velddrif, and Mediter-

ranean populations, especially during the younger stages. With age, individuals

may become distorted and display more irregular growth. It is of interest to note

that of the 15 Klein Brak specimens, 209 x 104 mm is the maximum size

attained, whereas five of the six Velddrif specimens, and four of the five Mediter-

ranean specimens exceed these dimensions, the Mediterranean specimens being,

on the average, the largest.

The variety rugosa mentioned by Lamy (1925) from the Atlantic coast of

Portugal and Morocco, would seem merely to be individuals showing a higher

degree of rugosity, several of the Klein Brak specimens being equally rugose.

The more pronounced roughness may well be due to the inhabiting of a substrate

more prone to shifting by greater water movement, as, for example, on the

exposed Atlantic coasts.

P. dreyeri, described by Van Hoepen (1940) from the Klein Brak deposits,

was separated, as already pointed out by Barnard (1964), on the basis of indi-

vidual variation.

The Plio-Pleistocene of the Mediterranean has yielded numerous examples

of P. glycymeris. The fossil form was originally designated as P. faujasi, but has

long been recognized as being synonymous with the living Mediterranean form

(Priolo 1966).

It would seem from the foregoing data, that the same specific name should

be applied to the living Mediterranean and West African forms, as well as to the

fossils from the Mediterranean, South West Africa and the Cape.

BRIEF ECOLOGICAL OBSERVATIONS

Ekman (1953, 1967), in discussing the fauna of the Mediterranean,

regarded Panopea glycymeris as an endemic form. The inaccuracy of this state-

ment was hinted at by Barnard (1962, 1964) and is now confirmed. Discussing

the hydrography of the Mediterranean, Ekman (1967: 81) states that the south-

east Mediterranean in summer has a temperature of 25—27°C, the rest having a

temperature of 20-25°C. In spite of the fact that winter temperatures drop well

below these levels, the area may nevertheless be regarded as warm-temperate.

The port of Dakar, situated on the bulge of Africa, lies within an area where

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 211

Fig. 8. Panopea glycymeris: Hinge lines.

a Mediterranean; b. Sicily; c. Portugal; d. Mocdmedes; e. Velddrif ; f. Klein Brak River.

212 ANNALS OF THE SOUTH AFRICAN MUSEUM

the surface water temperature is usually above 20°C, although at 100 m depth

the annual mean temperature is 18° or 19°C (Ekman 1967: 56).

In contrast to the present-day cold-temperate conditions of the sea off the

west coast of southern Africa, and the only slightly warmer conditions off the

south coast, Plio-Pleistocene conditions must have been rather different. It has

been suggested that along both the west and south coasts of South Africa, warmer

water than that of the present supported a characteristically warm-water fauna

(Carrington & Kensley 1969; Visser & Schoch 1973) during part of the Pleistocene.

The presence of Panopea in the Plio-Pleistocene deposits is yet another indication

of warmer conditions in the past. With the change to the present cold-water

regime, Panopea apparently died out at the southern end of its range, probably

due to the inability of either the juvenile forms or the adults to adapt physio-

logically to the changed conditions. It is of interest to note that amongst the

sediments removed from between valves of Panopea from the Velddrif area,

valves of the small bivalve Nuculana bicuspidata were found. This species, which

has also been recorded from the marine terraces of the Cape Cross area, has a

present-day distribution from Angola to Mauritania (Nicklés 1950). The sedi-

ments from between valves of the Klein Brak deposits yield, amongst a large

assortment of molluscs, numerous examples of the small trochid gastropod

Cantharidus fultoni (now only occurring alive from Mocambique northwards

along the East African coast), and a species of the bivalve Diplodonta provi-

sionally recorded by Barnard (1962) as senegalensis, but which would appear to

be very similar to D. diaphana, recorded from the Quaternary of Benguela,

Angola, and living from Angola to Mauritania. These isolated examples would

all seem to point to the existence of warmer water conditions along the west and

south coast of South Africa during the Pleistocene.

SUMMARY

The Plio-Pleistocene specimens of Panopea from South Africa are compared

with shells of living material from the Mediterranean and West Africa. As no

differences between the various specimens could be found, it is decided that all

belong to the same species, viz. Panopea (Panopea) glycymeris (Born).

ACKNOWLEDGEMENTS

I should like to thank the following individuals for making the loan of

material possible: Mrs K. M. Way of the British Museum (Natural History);

Dr M. de L. Paes da Franca of Lisbon; Mlle A.-M. Testud of the Muséum

National d’Histoire Naturelle, Paris; the Director of the National Museum,

Bloemfontein; Drs P. J. Roussouw and J. N. Theron of Geological Survey;

Mr A. Tankard and Mr M. R. Cooper of the South African Museum for useful

criticism.

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 2]3

REFERENCES

Apams, H. & Apams, A. 1853-58. The genera of recent Mollusca. 1-3. London: Van Voorst.

BARNARD, K. H. 1951. A beginner’s guide to South African shells. Cape Town: Maskew Miller.

BARNARD, K. H. 1962. Revised list of South African late Tertiary and Pleistocene marine

Mollusca. Trans. R. Soc. S. Afr. 36: 179-196.

BARNARD, K. H. 1964. Contributions to the knowledge of the South African marine Mollusca.

Part V. Lamellibranchiata. Ann. S. Afr. Mus. 47: 361-593.

BLAINVILLE, H. M. D. DE, 1825-27. Manuel de malacologie et conchologie. Paris: Levrault.

Born, I. von. 1778. Index rerum naturalium Musei Caesarei Vindobonensis. Pars I. Testacea.

Vindobonae: Krauss.

Born, I. von. 1780. Testacea Musei Caesarei Vindobonensis. Vindobonae: Mayer.

Bosc, L. A. G. 1802. Histoire naturelle des coquilles. 3. Paris: Roret.

Brown, T. 1827. I/lustrations of the recent conchology of Great Britain and Ireland. Edinburgh:

Lizars.

Brown, T. 1844. Illustrations of the recent conchology of Great Britain and Ireland. 2nd ed.

London: Smith, Elder.

CARRINGTON, A. J. & KENSLEY, B. F. 1969. Pleistocene molluscs from the Namaqualand coast.

Ann. S. Afr. Mus. 52: 189-223.

CHEMNITZ, J. H. 1780-95. Neues sytematisches Conchylien-Cabinet. Niirnberg: Raspe.

CHENU, J. C. 1862. Manuel de conchyliologie et de paléontologie conchyliologique. Paris:

Masson.

CuiEsA, C. 1932. Fossili miocenici dell’Isola di Coo (Egeox). Boll. Soc. geol. ital. 51: 171-188.

CLESSIN, S. 1895. Gastrochaenidae [including Saxicavidae]. In: MARTINI-CHEMNITZ. Syste-

matisches Conchylien-Cabinet. 2nd ed. 11(4a). Niirnberg: Bauer & Raspe.

Crosse, H. 1884. [Panopaea glycimeris Born.] J. Conch., Paris 32: 191. (Not seen.)

DAUTZENBERG, P. 1910. Contribution a la faune malacologique de l’Afrique occidentale. Act.

Soc. linn. Bordeaux 64: 1-174.

DAUTZENBERG, P. 1912. Mollusques marins. Mission Gruvel sur la céte occidentale d’Afrique

(1909-1910). Annls Inst. océanogr., Monaco § (3): 1-111.

DEBEAUX, O. 1884. [Panopaea Aldrovandi.] Naturaliste 6: 510. (Not seen.)

De Grecorio, A. 1929. Iconografia conchiologica Mediterranea. 4. Panopaea glycemeris Born.

Pan-norvegica Spengl. e Mya truncata G. Annls Géol.Paléont. 51: 1-21.

DEsHAYES, G. P. 1832. Histoire naturelle des vers . . .3. In: Encyclopédie méthodique. Paris.

DesHAYES, G. P. 1835. Histoire des mollusques. In: LAMARCK, J. B. P. Histoire naturelle des

animaux sans vertébres. 2nd ed. 6. Paris: Bailliére.

Desuayes, G. P. 1843. Traité élémentaire de conchyliologie, avec les applications de cette

science a la géologie. I. Paris.

De STEFANI, T. 1941. Fauna malacologica di un nuovo giacimento fossilifero del ‘Piano

Siciliano’ in Palermo. Boll. Soc. geol. ital. 60: 275-281.

EKMAN, S. 1967. Zoogeography of the sea. London: Sidgwick & Jackson.

FISCHER, P. 1880-1887. Manuel de conchyliologie et de paléontologie conchyliologique. Paris:

Savy.

Forses, E. & HANLEY, S. C. T. 1853. A history of British Mollusca and their shells. 1. London:

Von Voorst.

FRANCA, M. L. P. DA. 1960. Contribuigao para o conhecimento da fauna malacoldgica de

Angola. (Terceira nota.) Mems Jta Invest. Ultramar (2) 15: 7-40. é

GHISOTTI, F. & STEINMANN, E. 1969. Panopea glycymeris (Born 1778). In: GuHisorTTI, F., ed.

Schede malacologiche del Mediterraneo: Sheet No. 78—Ac-—01. Como: Societa Mala-

cologica Italiana.

GMELIN, C. 1790. C. a Linné . . . Systema naturae. Ed. 13. Lipsiae: Beer.

Gray, J. E. 1847. A list of the genera of recent Mollusca, their synonyma and types. Proc. zool.

Soc. Lond. 1847: 129-219.

HANLEY, S. C. T. 1842-56. An illustrated and descriptive catalogue of recent bivalve shells,

forming an appendix to the Index testaceologicus. London: Williams & Norgate.

HOoeEPEN, E. C. N. VAN. 1940. Oor nuwe skulpe uit ou strandwalle en oor die ouderdom van die

afsetting. Tydskr. Wet. Kuns. (n.r.) 1: 185-194.

IMPERATORI, L. 1961. Documentos para el estudio del Cuaternaria alicantino. Estudios geol.

Inst. Invest. geol. Lucas Mallada 13: 141-152.

214 ANNALS OF THE SOUTH AFRICAN MUSEUM

LAMARCK, J. B. P. A. DE M De. 1818. Histoire naturelle des animaux sans vertébres. 5. Paris:

Bailliére.

Lamy, E. 1925. Révision des Saxicavidae vivants du Muséum National d’Histoire Naturelle de

Paris. J. Conch., Paris 68: 261-283.

LocarD, A. 1886. Prodrome de malacologie francaise. 2. Mollusques marins. Lyon.

LocarD, A. 1892. Les coquilles marines des cétes de France. Paris.

Ma tatesTA, A. & Nicosia, M. L. 1955. Contributi alla conoscenza della faune Neogeniche e

Quaternarie della Sicilia. I. Boll. Serv. geol. Ital. 77: 173-180.

Mayer, C. 1870. Catalogue systématique et descriptif des mollusques tertiaires du Musée

Fédéral de Zurich. IV. Vjschr. naturf. Ges. Ziirich 15: 31-82.

MENARD DE LA GROyE, F. J. B. 1807. Sur un nouveau genre de coquille de la famille de

Solénoides (Panopea). Annls Mus. Hist. nat. Paris 9: 131-139.

MENARD DE LA GROYE, F. J. B. 1808. Note sur le Panopea Aldrovandi. Annls Mus. Hist. nat.

Paris 12: 464465.

MontTacu, G. 1808. Testacea Britannica. Suppl. London.

MOoNTEROSATO, T. A. DiI. 1875. Nuova rivista delle conchiglie Mediterranee. Palermo. Atti

Accad. Sci. Lett., Palermo (n.s.) 5: 1-50.

MONTEROSATO, T. A. DI. 1878. Enumerazione e sinonimia delle conchiglie Mediterranee.

Palermo. G. Sci. nat. econ. Palermo 13: 61-115.

MONTEROSATO, T. A. DI. 1889. Coquilles marines marocaines. J. Conch., Paris 37: 20-40,

112-121.

Moorg, C. R. 1969. Treatise on invertebrate paleontology. Part N. Vol. 2. Mollusca 6: Bivalvia.

Kansas: Geological Society of America & University of Kansas.

NICKLES, M. 1950. Mollusques testacés marins de la céte occidentale d’ Afrique. Paris: Leche-

valier.

Nosre, A. 1940. Fauna malacologica de Portugal. 1. Molluscos marinhos e das aquas Salobras

de Portugal. Barcelos: Minho.

PALLARY, P. 1900. Coquilles marines du littoral du Departement d’Oran. J. Conch., Paris 48:

211-422.

PALLARY, P. 1920. Exploration scientifique du Maroc. Mission zoologique. Malacologie (1912).

Explor. scient. Maroc. 2: 1-108.

PETERSEN, C. G. J. 1888. Om de skalbaerende molluskers udbredningsforhold i de danske have

indenfor Skagen. Copenhagen. (Inaug.—dissertation.)

PETIT DE LA SAUSSAYE, S. 1869. Catalogue des mollusques testacés des mers d’ Europe. Paris:

Savy.

PHILIPPI, R. A. 1836. Enumeratio molluscorum Siciliae. 1. Berolini: Schropp.

PHILLIps, J. W. 1970. Geoduck hunting. Sea Front. 16: 246-250.

PoTiEz, V. L. V. & MICHAUD, A. L. G. 1844. Galerie des mollusques, ou Catalogue méthodique,

descriptif et raisonné de mollusques et coquilles du Muséum de Donai. Paris: Bailliére.

PRIOLO, O. 1966. Sulla Panopea glycymeris Born vivente e fossile. Boll. Sed. Accad. gioenia Sci.

nat. (4) 8: 669-678.

REEVE, L. A. 1841. Conchologia sytematica. 1. London: Longman.

Rocers, A. W. 1906. A raised beach deposit near Klein Brak River. Rep. geol. Commn Cape

Good Hope 10: 293-296.

RuGGIERI, G. 1943-44. I] Calabriano e il Siciliano nella Valle del Santerno (Imola). G. Geol. (2)

17: 95-113.

RuGGIERI, G. 1948. La malacofauna del Calabriana Romagnolo. G. Geol. (3) 20: 63-110.

RuGGIERI, G. 1949. Contributi alla conoscenza della malacofauna e della stratigrafia del

Pliocene e del Quaternario. G. Geol. 21: 86.

SCHWARZ, E. H. L. 1910. The Pleistocene deposits of Port Elizabeth. Trans. geol. Soc. S. Afr.

12: 112-118.

SmiTH, E. A. 1903. A list of species of Mollusca from South Africa, forming an appendix to

G. B. Sowerby’s ‘Marine shells of South Africa’. Proc. malac. Soc. Lond. 5: 354-402.

SmiTH, E. A. 1906. Jn: ROGERS, A. W. A raised beach deposit near Klein Brak River. Rep. geol.

Commn Cape Good Hope 10: 293-294.

SOwERBY, G. B. 1873. Monograph of the genus Panopaea. In: REEVE, L. A. Conchologia iconica.

19. London: Reeve.

Sowerby, G. B. 1889. Some further notes on marine shells collected at Port Elizabeth, South

Africa, with descriptions of new species. J. Conch., Lond. 6: 6-15.

THE STATUS OF THE PLIO-PLEISTOCENE PANOPEA IN SOUTHERN AFRICA 215

Stow, G. W. 1871. On some points in South African geology. Part 1. Q. JI geol. Soc. Lond. 27:

497-522.

THIELE, J. 1935. Handbuch der systematischen Weichtierkunde. 2. Jena: Fischer.

TRYON, G. W. 1869. Catalogue of the families Saxicavidae, Myidae and Corbulidae. Am. J.

Conch. 4: 59-68.

TurToN, W. 1819. A conchological dictionary of the British Islands. London: Booth.

TurTON, W. 1822. Conchylia Insularum Britannicarum: 42. Exeter.

VALENCIENNES, A. 1839. Description de l’animal de la Panopée australe et recherches sur les

autres espéces vivantes ou fossiles de ce genre. Archs Mus. Hist. nat. Paris 1: 1-38.

VALENCIENNES, A. 1843. Panopaea. In: CHENU, J. C. Illustrations conchyliologiques. Paris:

Fortin, Masson.

Visser, H. N. & ScHocu, A. E. 1973. The geology and mineral resources of the Saldanha Bay

area. Mem. geol. Surv. S. Afr. 63: i-iv, 1-150.

WINCKWoRTH, H. C. & WINCKWORTH, R. 1935. Eastonia rugosa and other Portuguese Mol-

lusca. J. Conch., Lond. 20: 162-163.

Woop, S. V. 1853. A monograph of the Crag Mollusca, with descriptions of shells from the Upper

Tertiaries of the British Isles. 2. Bivalves: 283. London: Palaeontographical Society.

Woop, W. 1825. Index testaceologicus. London: Wood.

Woopwapb, S. P. 1855. On Panopaea Aldrovandi, Lam. Proc. zool. Soc. Lond. 1855: 218-221.

ZeI, M. 1960. [Panopaea glycimeris.| Boll. Soc. Nat. Napoli 69: 184. (Not seen.)

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FIscHEerR, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FiscHER, P.-H., DUVAL, M. & RarFry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konun, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 19605. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen -und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

Brian Kensley

THE-STATUS OF THE PLIO-PLEISTO@GERE

PANOPEA IN SOUTHERN AFRICA

(MOLLUSCA, BIVALVIA, HIATELLIDAE)

VOLUME 65 PART 8

se 7 Ct

OF ee KO) Ae

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 Band

July 1974 Julie

Part $ Deel

PETROLOGY AND ORIGIN OF THE PHOSPHORITE

AND ALUMINIUM PHOSPHATE ROCK OF THE

LANGEBAANWEG—SALDANHA AREA,

SOUTH-WESTERN CAPE PROVINCE

ANTHONY J. TANKARD

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 2038S E28), 322, Soa) 52) 5-7-9):

6(1, t.-p.i.), 70-3), 8, 9(1-2), 10(1),

G25) 7, tps 240) Om 3103), 33

Price of this part/Prys van hierdie deel

R5,20

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 47 X

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

PETROLOGY AND ORIGIN OF THE PHOSPHORITE AND

ALUMINIUM PHOSPHATE ROCK OF THE LANGEBAANWEG-—

SALDANHA AREA, SOUTH-WESTERN CAPE PROVINCE

ANTHONY J. TANKARD

South African Museum, Cape Town

(With 25 figures and 3 tables)

[Ms. accepted 8 August 1973]

CONTENTS

PAGE

Introduction ‘ ; : . : : ‘ eT,

Methods. ; ee Sc ies : : , e219

Geologic setting . : : : : Al)

Petrology of the phosphorites . Paige : : ae 220)

Pelletal phosphorite . 2 ; : : te ao

Phosphate rock . : 5 : : : es:

Mineralization . AS i)

Chemical composition of the phosphorite 0) Dag

Aluminium phosphates. z : : 4 = 240

Genesis of the phosphate deposits . ‘ ; s 2 2A2

Phosphorite. : : ‘ : ; . 242

Aluminium phosphate ; 2 : : . 246

Summary . : é : : : . a 247.

Acknowledgements : , : d ; ; . 248

References) =) 1). Te ee : : : . 248

INTRODUCTION

Phosphorite is the term used to describe marine sedimentary deposits which

contain more than 18 per cent P,O;, approximating to 50 per cent apatite

(Bushinsky 1966). The apatite, generally a carbonate-fluorapatite (Altschuler et

al. 1958), forms authigenically in a shallow marine environment, giving rise to

phosphate rock and nodules. Reworking of the phosphate rock produces pelletal

phosphorite. Since the pelletal phosphorite is a product of erosion of a phos-

phatic rock it must necessarily be a shallow-water deposit whereas nodules and

tock phosphate can be found in considerably deeper water. The present-day

occurrence of phosphorite on land may be attributed to both tectonic and -

eustatic causes.

Although submarine phosphorites were first recorded on the Agulhas Bank

by the Challenger Expedition of 1873-6 (Murray & Renard 1891), relatively few

studies have been made of South African phosphorites, viz. Haughton (1932),

Cayeux (1934), Frankel (1943), Parker (1971), Parker & Siesser (1972), Parker &

Simpson (1972), Summerhayes et ai. (1972), Summerhayes (1973), Tankard

(1974). However, a voluminous literature discusses phosphorite deposits from

other parts of the world (for example, Sheldon 1964; D’Anglejan 1967; McKelvey

1967; Rooney & Kerr 1967; Tooms et al. 1969).

aly,

Ann. S. Afr. Mus. 65 (8), 1974: 217-249, 25 figs, 3 tables.

218 ANNALS OF THE SOUTH AFRICAN MUSEUM

IBE

St Helena

Bay

Berg

JSON

2 Paternoster

\e/

33'S ‘

oe ee

Vredenburg

I) Baard's Quarry

33S

N x

as Groot Springfonte!

Saldanha Bay aaa borehole

jutten © SS >* 5 pLangebaan

xeon Bay

0 10 So

eeeemne———— {eee ——— ome!

: KILOMETERS . U

—_—

Fig. 1

Locality map of farms prospected. 1, Pelgrimsrus; 2, Noodhulp; 3, Besterskraal; 4, Sand-

heuwel: 5, Langlaagte; 6, Witteklip; 7, Langeberg; 8, Muishondsfontein; 9, Tiekosklip;

10, Waschklip; 11, Duyker Eiland.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 219

On the granite hills north and south of Saldanha Bay there are aluminium

phosphates which have been discussed by Du Toit (1917), Hutchinson (1950)

and Visser & Schoch (1973). They probably originated by leaching of guano

deposits by ground water and the subsequent alteration of the granite bedrock

to form aluminium and aluminium iron phosphates.

METHODS

All sediment samples were obtained from Chemfos Ltd who used a Selby

coring technique, driving a 1,07 m long cylinder into the sediment. The coring

was part of their mining and prospecting programme.

Pelletal phosphorite concentrates were obtained by heavy liquid (bromo-

form) separation from the detrital quartz.

Scanning electron photomicrographs were obtained on the JEOL JSM U3

instrument at Rhodes University. Specimens were coated with a gold-palladium

alloy, vacuum evaporated. Coating was done from high and low angles and the

samples rotated to obtain an even coating.

GEOLOGIC SETTING

The phosphorite deposits with which this report is concerned are situated

primarily about the perimeter of an erosional basin between Langebaanweg and

Saldanha (Fig. 1). The areal extent of the Varswater Formation has been

established by examination of several hundred logs of boreholes, both within

this basin area and on the Atlantic coast between Saldanha Bay and St. Helena

Bay. The stratigraphy and lithology of the Varswater Formation and associated

sediments are summarized in Table 1. They have been fully discussed elsewhere

(Tankard in press).

Lying unconformably below the Varswater Formation on Langeberg is the

Middle Miocene basal bed (Tankard in press), a phosphatic sandstone in

which the apatite has formed authigenically. This phase of deposition was fol-

lowed by regression and tilting. In the Pliocene a transgressing sea pushed

deltaic marsh sediments ahead of it until it reached a temporary stillstand; the

littoral zone is now at 30 m above sea-level. These temporarily stable conditions

allowed a barrier bar to build up, behind which estuarine conditions prevailed.

The estuary was fed by a river from the north-east. The final transgression

reworked the older deposits. The maximum extent of the final transgression is

now recorded at 50 to 55 maz.s.l. Erosion of the basal bed during the final trans-

gression supplied the pelletal phosphorite which characterizes the Pelletal

Phosphorite Member.

There is little evidence in these sediments of higher than present sea-levels

during the Pleistocene. The Pleistocene is represented only by sheets of wind-

blown sand blanketing the earlier deposits, while a system of shallow streams

flowed off the northern slope of Anyskop, a hill on Langeberg, drawing phos-

phate from the Varswater Formation.

220 ANNALS OF THE SOUTH AFRICAN MUSEUM

Stratigraphic Unit

Pleistocene aeolian sands

River channel sediments

Quaternary

| Pelletal

| Phosphorite

Member

Pliocene

Beach Gravel

Member

Varswater Formation

Kaolinitic clay

(freshwater)

Quartzose sand

Fluvial Sand

unknown origin

| Middle Miocene |

Basal bed

Mottled silty—clay

TABLE 1

Stratigraphy of the Varswater Formation and associated sediments

Maximum

thickness

41m

0,5—2 m

Lithology

Calcareous—quartzose sands, medium- to

fine-grained, moderately sorted

Greenish-white clayey sands with channels of

clayey sand and phosphatic sandstone

Upper and lower boundaries defined as 2%

P.O; cutoff. Moderately sorted, medium to

fine phosphatic—quartzose sands. Phos-

phate present as sub-spherical pelletal

phosphorite. Lenses and concretions of

phosphatic sandstone

(i) Estuarine facies: yellowish-brown sandy

silt, poorly sorted. 30-40% mud. Little

phosphate. Mammal fossils abundant

(ii) Fluvial facies: coarse sand intercalated

with fine sand. Moderately sorted

quartzose sands

Consolidated quartzose sands, frequently

phosphate mineralized. Well-rounded

beach cobbles and gravelly-sands. Shell

casts of molluscs of warm-water affinity

Grey-black carbonaceous clay, pyrite, grass

and Podocarpus pollens

Fine quartzose sand

UNCONFORMITY-TILTING

0,5-1,5 m | Generally poorly sorted, fine-grained phos-

phatic—quartzose sandstone. Hoedjiespunt:

massive lens of bedded apatite, less than

1% quartz

Silty-clay, poorly sorted, iron stained. No

phosphate. Extent unknown

PETROLOGY OF THE PHOSPHORITES

A detailed optical study of the phosphorite pellets and phosphatic sandstone

is largely precluded because of the submicroscopic size (0,25—4 ) of the apatite

mineral and the admixture of argillaceous, carbonaceous and ferruginous

impurities. The pellets may all be classified as true phosphorite, having a P,O;

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK pps

content in excess of 18 per cent. But this does not necessarily hold true for all the

phosphatic sandstone specimens, which will be described along with the true

phosphorite.

PELLETAL PHOSPHORITE

The term pellet as used throughout this report is not used in a genetic sense

or even restricted to particles smaller than 0,15 mm (2,740) as defined by Folk

(1962). ‘Pellet’ is used here entirely in a descriptive sense. The mean grain-size of

the pellets is coarser than that of the associated quartz grains in the sediment.

The mean pellet-size on Langeberg is 2,020 (range 1,8-2,30); on Witteklip

they are finer grained, 2,270 (range 1,5-2,70). North of Hondeklipbaai the

pellets are typically coarser (0,5—1,50). The amount of pelletal phosphorite in

silt-size grades is negligible. The pelletal particles are present in three distinct

forms. Those of biogenous origin are present only in trace amounts. The

inorganic pellets appear to owe their gross morphology to particle size. In the

coarser fractions (2,0 to —0,59) the pellets are of platey or irregular appearance

while below 2,09 ovoidal structureless pellets predominate.

The increase in degree of roundness with diminishing size displayed by the

pelletal phosphorite is opposite to the trend of the quartz grains in the same

sediment where roundness improves with increasing grain-size. This reversal of

behaviour of the pellets is readily explained by their soft (hardness of apatite is

4,5 to 5) but brittle nature. The larger particles tend to fracture and remain

subangular. Finer than 20, the phosphorite abrades more rapidly than the

quartz. The spheroidal shape of the pellets results in their behaving differently

from the coarser plates under turbulent conditions. The sub-spherical pellets are

probably in hydraulic equilibrium with the coarser platey pellets. The high degree

of rounding and sorting of the pellets is characteristic of deposition in a littoral

environment.

Generally the edges of the coarser platey pellets show some signs of wear,

while many of the grains have a conchoidal fracture and often a striated surface.

Undoubtedly many of these grains are actually phosphatized mollusc shell

fragments, but such an origin does not account for all of them as D’Anglejan

(1967) suggested for his phosphorites. The pellets range in colour from pale

yellow to orange to deep red. They are usually of fresh appearance and often

translucent. Frequently they have white blemishes due to finely divided clay

material, or are mottled black by organic carbon. Generally these grains are of

lighter and more yellow coloration than the smaller ovoidal pellets. Frankel

(1943) analysed the white portion of nodules from Langebaanweg as mont-

morillonite.

The ovoidal pellets constitute more than 90 per cent of all pellets in the

2,5—3,59 fraction. They are regular, sub-spherical in shape (Fig. 2). Rod-shaped,

but nevertheless well-rounded pellets are also found. Rooney & Kerr (1967) have

described similar particles with a groove running down the length of the rod,

which they thought were probably minute bones, although they concede the

Doe ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 2

Morphology of typical pelletal phosphorite.

possibility of coprolites in their sediments. Arakawa (1971) illustrates some

faecal pellets of invertebrates which are very similar to those found in the Vars-

water Formation. Very rare are some well-rounded pellets composed of aggre-

gates of smaller pellets and phosphate cemented. The ovoidal pellets are usually

brown, orange or red, but black and green pellets have also been noted. Towards

the base of borehole W’5 (Witteklip) the pellets are mostly black, due to large

amounts of organic carbon (Fig. 6). Black mottling and white blemishes are

common in most of the ovoidal pellets.

Included with the biogenous particles are translucent and opaque phos-

phatized echinoid spines, foraminifers, minute fish teeth, Bryozoa and coprolites.

The echinoids spines generally have rounded extremities and the foraminifers

(predominantly Elphidium sp.) also show considerable wear. The coprolites are

cylindrical and usually have a groove down their length.

Because of the cryptocrystalline state of the apatite, the pellets are difficult

to study in thin-section. And the poorly crystalline state results in low intensity

X-ray diffractometry peaks. In thin-section the pellets are yellowish brown and

isotropic to very slightly anisotropic. Some pellets have an outer rim of aniso-

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 223

Fig. 3

Structure of pellet that has resulted from the migration of impurities away from the rim.

tropic apatite. They are usually structureless, although a structure may occur

after apatite has grown about an older pellet, or the growth of the apatite

crystals has pushed aside the carbon and argillaceous inclusions (Fig. 3).

Frequently the pellets contain subangular silt-size quartz grains randomly

scattered throughout the groundmass (Fig. 4). These frequently give the impres-

sion of a nucleus. But the presence of organic carbon and argillaceous material

within pellets suggests precipitation within the pore-spaces of the sediment, the

silt-size quartz being also present in these pore-spaces. (Attempts to determine

the organic carbon content by direct combustion, first in an atmosphere of

nitrogen, and again in an atmosphere of oxygen, produced inconclusive results.

However, hydrochloric acid leaching of the apatite left behind a black fibrous

organic carbon-like residue which was rapidly oxidized by peroxide.) In only one

partially corroded pellet was a well-developed oolitic structure present, showing .

concentric shells of radiating anisotropic francolite. Figure 4 shows the occur-

rence of a well-rounded grain that is predominantly quartz but includes one

portion of phosphorite. Although the phosphorite formed originally as an

authigenic mineral within the pore-spaces of an older formation, in the latest

cycle of erosion the phosphorite has become associated with the detrital compo-

nents of the sediment and has been abraded to the typical pelletal form. Figure 5

demonstrates typical pellet form. It also demonstrates the ubiquitous nature of

organic carbon in the pellets, and shows a pellet of biogenous origin. At the base

of W’5 (Witteklip) the pellets are saturated with organic carbon (Fig. 6).

224 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 4

Pellets with quartz grain inclusions. Arrow indicates a well-rounded grain composed of part

phosphorite, part quartz.

Fig. 5

Typical structure of pelletal phosphorite. Arrow indicates biogenic particle.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 225

Fig. 6

Pellets coloured black due to carbonaceous material.

= Shane

Organic material present as small laths.

226 ANNALS OF THE SOUTH AFRICAN MUSEUM

Many of the pellets contain small laths (Fig. 7), possibly of organic material,

but it is impossible to identify them.

Phosphorite pellets with oxidized rims are common. Similar features led

Rooney & Kerr (1967) to suspect extensive reworking. The ovoidal pellets have

had a longer period for oxidation of the organic constituents to take place,

whereas the more angular pellets have a fresher appearance due to their com-

parative youthfulness. The pellets owe their colour to disseminated aggregates

of carbonaceous, argillaceous and ferruginous impurities. The iron oxide (Fe,O,)

content of many pellets is as high as 3,4 per cent. Opaline silica is frequently

present within the pellets, and in other cases has formed a shell about the pellet

from within which the apatite has been dissolved, leaving a fragile empty shell.

(D’Anglejan (1967) records a similar phenomenon.) The coarser platey pellets

have in many cases a texture typical of shell debris. Some sections strongly

resemble young bone or cartilage (Fig. 8) and tooth dentine (Fig. 9). It was

1304

ONG /)

1g.

Pellet structure mentee ce bone or cartilage.

found that some of the pellets on Witteklip have a texture very similar to that of

the Hoedjiespunt phosphorite. But chemical analyses show the Hoedjiespunt

phosphorite to be dahllite and the Witteklip pellets predominantly francolite.

The Langeberg pellets likewise have a texture very similar to that of the matrix

of the local basal bed.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK OS |

Fig. 9

Structure similar to tooth dentine.

Scanning electron photomicrograph showing apatite crystals in phosphorite pellet

228 ANNALS OF THE SOUTH AFRICAN MUSEUM

Figure 10 shows the minute dimensions of the apatite crystallites in the phos-

phorite. The texture of the fracture surface of the pellets is further illustrated in

Figures 11 and 12.

PHOSPHATE ROCK

Phosphatic sandstones have resulted from the precipitation of collophane

(a useful term that encompasses all the cryptocrystalline carbonate apatites)

within the pore-spaces of a quartzose sand. At Hoedjiespunt the phosphorite

horizon, 1,5 m thick, contains less than 1 per cent quartz. Bushinsky (1966)

defines phosphorite as a rock containing more than 18 per cent P,O, (approxi-

mating to 50 per cent apatite). In only 38,1 per cent of the phosphatic rocks of this

study did the P,O,; concentration exceed 18 per cent. The highest P,O, value is

27,2 per cent. The average value for the basal bed is 14,9 per cent P,O,. Parker

& Siesser (1972) report an average of 15 per cent P,O, in the continental margin

phosphate rocks.

1. Miocene basal bed

The rock is brown with a surface that is polished and is undulating or pitted

by differential erosion or burrowed by marine animals. Essentially the rock of

Fig. 11

Scanning electron photomicrograph showing structure of pellet fracture surface.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 229

Fig. 12

Scanning electron photomicrograph showing typical structure of pellet fracture surface.

this horizon consists of fine sand-size quartz embedded in a matrix of finely

divided argillaceous and organic material that is collophane cemented. Frankel

(1943) has identified the argillaceous material as montmorillonite. The detrital

quartz fraction constitutes from 20 to 95 per cent of the rock. It has a mixed grain

population, generally poorly sorted, and frequently shows considerable iron

staining. Fe,O, percentage ranges from 0,3 to 6,5.

Near the south-western corner of the New Varswater Quarry the basal bed

is markedly conglomeratic, and in places brecciated (Fig. 13). Fractures in the -

larger non-phosphatic sandstone inclusions are penetrated by collophane. These

inclusions are usually rounded, while some have been almost completely

phosphate mineralized. Figure 14 illustrates a rather problematical occurrence of

phosphate rock. Here phosphate-rich solutions have differentially impregnated

the host rock. There are no textural differences across the phosphate/phosphate-

free interfaces. Most of these phosphate rocks may be described as medium- to

fine-grained collophane packstones. In the majority of cases the collophane has

undoubtedly originated by direct precipitation of phosphate within the pore-

spaces of a quartz sand from phosphate-rich waters. But in a few cases precipita-

230 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 13

Fig. 14

Preferential phosphate mineralization of basal bed.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 2351

tion of the phosphate has possibly been induced by algae present in the sediment.

Figure 15 shows a structure that has developed in the form of concentric shells

of different mineral composition. In Figure 16, which is an enlargement of part

of the previous figure, it is shown that this structure has developed independently

of the basic sediment texture. The concentric shells reflect zones of enrichment by

collophane and sometimes iron oxide. Iron oxide ( ?goethite) would appear to be

ubiquitous in these phosphate rocks, but it could be younger than the phosphate

component in many cases.

Heavy minerals, mainly ilmenite, are present only in trace amounts. Bone

fragments, which are common, are always completely phosphatized. Shell

debris is much less common.

As the amount of quartz varies considerably so does the number of inter-

grain contacts. Grain contacts vary from nil to three per grain. A grain contact of

one indicates an originally loosely packed sediment, while a grain contact of

three per grain suggests close packing. The quartz fraction is a mixed grain

Fig. 15

Concentric structure of phosphorite.

P2202 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 16

Enlarged section of previous plate showing that concentric structure has developed indepen-

dently of sediment texture.

Fig. 17

Phosphorite pellets containing quartz grains in phosphatic sandstone of basal bed.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 233

population, the largest grain-size being about 0,4 mm and well rounded. Many

of these well-rounded grains have been broken in the last cycle of erosion, the

rounded aspect having been inherited from a previous cycle. Corrosion of

quartz grains has also occurred. An interesting feature of the phosphate rock is

the frequency of pellets of phosphorite with much silt-size quartz in the matrix

(Fig. 17). Some of the included phosphorite pellets are, like the quartz, corroded.

Sometimes pore-spaces are lined with drusy quartz. Generally the matrix is

yellow brown under plane-polarized light. It consists often of a rim of anisotropic

francolite grown from the walls of the voids, the remaining space being filled

with isotropic collophane. The structure suggests precipitation of the apatite in

the pore-spaces of the sediment. The drusy francolite is illustrated in Figure 18.

Only very seldom was any calcite found.

As the beach cobbles from the Beach Gravel Member have been derived by

erosion of the basal bed, there is little that need be said about them. Thin-sections

showed a composition of about 90 per cent by volume of detrital quartz, which

has a close-packed structure (grain contact of 3 per grain). The matrix, as for the

basal bed, is composed of isotropic or slightly anisotropic collophane.

The apatite of Hoedjiespunt has a very different aspect. The rock contains

3,6 per cent SiO,; 1,8 per cent Al,O,; 0,3 per cent Fe,O, and 34,0 per cent P,O;.

Thin-section study and X-ray diffraction data confirm that little of the SiO, is

——

—~>

Fig. 18

Drusy francolite grown from walls of the voids in the phosphatic sandstone.

234 ANNALS OF THE SOUTH AFRICAN MUSEUM

present as free quartz. Even the clay mineral that accounts for the balance of the

SiO, and Al,O, is present in very small amounts. The low fluorine content defines

the mineral as dahllite. Like the francolite of Langeberg it is isotropic to very

slightly anisotropic under crossed nicols. Shell debris, bryozoan remains and

foraminifers are set in a collophane (dahllite)/micrite groundmass (Fig. 19). The

concentration of iron oxide is considerably lower than that of the Langeberg

basal bed. Planktonic foraminifers are present. The Hoedjiespunt phosphorite

is partly phosphatized micrite and partly phosphatized microcoquina. Figure 20

shows small cleavage flakes of phosphorite set in a collophane matrix, while

there has also been collophane precipitation along fractures within the initial

collophane matrix. An oolitic structure is shown in Figure 21. This type of

structure is common in the Hoedjiespunt phosphatized wackestone.

2. Phosphate rock of the Varswater Formation

Precipitation of isotropic collophane within the pore-spaces of the quartz

sands has been influenced by local concentrations of organic matter. This has

resulted in the development of thin lenses and concretions of phosphatic sand-

stone. Figure 22 shows a tortoise bone that has acted as a nucleus for phosphate

deposition which has followed the contours of the bone. Although the bone is

completely phosphatized, the maximum phosphate deposition outside the bone

Fig. 19

Hoedjiespunt phosphorite showing foraminifers and shell debris.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 235

Fig. 20

Fig. 21

Oolitic structure, Hoedjiespunt.

236

ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 22

Tortoise bone that has acted as a nucleating centre for phosphate precipitation.

Fig. 23

Typical phosphatic sandstone lens structure from Varswater Formation.

SRI

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 237,

is slightly separated from the bone itself. Figure 23 shows the type of local con-

centration of phosphate frequently encountered. The detrital components of the

phosphatic packstones are always similar to that of the surrounding sediment,

and they include also phosphorite pellets. Cementation has taken place by

precipitation of francolite about the grains to give a radially disposed crypto-

crystalline francolite with the pore-spaces filled with clear collophane or mud.

Only occasionally is drusy quartz found. In some of the packstones a crudely

graded bedding has been observed. The colour of these phosphatic packstones

ranges from pale brown to black. Samples from borehole S12 (42,8 m) were

analysed. The brown variety contained 11,6 per cent P,O, and the black variety

19,6 per cent P,O;. Towards the base of the Varswater Formation a collophane

mudstone is frequently encountered (Fig. 24). It contains 5 to 10 per cent detrital

quartz set in a fine phosphatized argillaceous matrix.

MINERALIZATION

Petrographic evidence suggests that at least three distinct periods of post-

depositional phosphatization have occurred. The conglomeratic phosphate rock

Fig. 24

Collophane mudstone.

238 ANNALS OF THE SOUTH AFRICAN MUSEUM

or packstone of the Langeberg basal bed shows two periods of mineralization.

The first phase is represented by fragments of an older phosphate rock as well as

phosphorite pellets. These older components were reworked and deposited to

form a new rock mass which has subsequently become phosphatized. In the basal

bed phosphate precipitation has resulted in a phosphate packstone only one

metre thick. Phosphate mineralization must have taken place fairly rapidly after

deposition. The presence of phosphatized cartilage or young bone material

would also suggest that very rapid phosphatization has occurred. Phosphatiza-

tion of the basal bed must have been nearly contemporaneous with sedimenta-

tion and must have depended upon the close proximity of upwelling of

phosphate-rich water (to be discussed later). Phosphatization appears to be

nearly uniform through much of the rock of the basal bed, although contact

with phosphate-rich water has increased the degree of phosphatization in a thin

layer at the surface. The overlying phosphatic packstone cobbles are identical

with the basal bed and have been formed from material from that bed, which

suggests that the phosphatization must have preceded the last phase of erosion.

On Langeberg phosphate mineralization has taken place within the pore-spaces

of a quartz sediment, mineralizing the argillaceous material already there.

Organic matter within the argillaceous material has probably behaved as a

catalyst. On Langeberg there is little evidence that this mineralization has pro-

ceded via a lime replacement mechanism. I have suggested that locally the phos-

phate precipitation has possibly been induced by algae. The algae would have

behaved in a similar way to the other organic matter. On Langeberg the phos-

phorite pellets of the Varswater Formation have a very similar aspect to that of

the matrix of the basal bed, although some do show concentric layering. In these

rare cases the concentric structure is probably the result of apatite growth about

an initial pellet. There is no evidence to suggest growth of the pellets about

nuclei, although silt-size quartz inclusions give the impression of nucleation.

Thin-section analyses show that the quartz was present in the matrix of the

original phosphorite, as was the organic material. Summerhayes (1970) argues

that if the francolite in the original phosphatic rock has grown from the walls of

the voids, these layers may act as lines of structural weakness. Disintegration

along such lines leads to the formation of pellets of collophane that contain

impurities but have a clear collophane margin.

On Langeberg the third phase of phosphate mineralization is marked by the

appearance of phosphatic sandstone lenses and concretions in the Varswater

Formation. The estuarine sediments are capped by a thin phosphatic sandstone

horizon. In the Pelletal Phosphorite Member the phosphatic sandstone lenses

and concretions include pelletal phosphorite in the same abundance as the

adjoining sediments. In this same member there are phosphatized foraminiferal

tests, indicating a lime replacement mechanism. Shell material, shark teeth,

bone fragments, echinoid spines and foraminiferal tests are all well

phosphatized.

On Hoedjiespunt the bedded apatite also demonstrates three phases of

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 239

phosphate mineralization. An original flaky phosphorite has been incorporated

into a later wackestone which has been phosphatized, and in a third phase of

mineralization fractures within this phosphatized wackestone have been the sites

of phosphate deposition. At Hoedjiespunt it would appear that the phosphorite

has resulted from collophane replacement of a micrite containing shell debris and

foraminiferal tests.

Within the Varswater Formation sedimentation appears to be confined to

deposition of fine material. Frequent erosional episodes have resulted in con-

centration of shark teeth and mollusc shells into definite horizons. Widespread

reworking is attested by the broken detrital grains and shell debris and worn

foraminiferal tests. All evidence suggests low rates of sedimentation.

CHEMICAL COMPOSITION OF THE PHOSPHORITE

The apatite mineral of all the pelletal phosphorites and the Langeberg

basal bed is francolite, while the apatite of the Hoedjiespunt phosphorite is

dahllite (Tankard 1974). Francolite is the name applied to an apatite con-

taining appreciable CO, and more than | per cent fluorine, whereas the name

dahllite has been applied to apatite containing abundant CO, but less than 1 per

cent fluorine (McConnell 1938).

Table 2 shows the chemical composition of pelletal phosphorite and

Hoedjiespunt phosphorite. The high SiO, content of the pelletal phosphorite

sample is deceptive in that it is a measure of the efficiency of the heavy liquid

separation. However, the results do give an idea of the chemical constitution of

the phosphorite. But much of the SiO, is bound up as a clay mineral where it

would be combined with the Al,O, and K,O. The Fe,O, content is high. Thin-

section examination shows that less than | per cent of the SiO, in the Hoedjies-

punt phosphorite is present as free quartz.

TABLE 2

Chemical composition of the phosphorite

Pelletal phosphorite Hoedjiespunt phosphorite

Ya Ya

SiO tg: oa 4,13 3,60

ANCOR Te ic 1,73 1,83

Fe,O; 7/7) 0,27

CaO 46,97 46,04

MgO 0,28 P29)

Na,O 0,68 1,27

KOO |. : 0,26 0,06

ELO , 2,43 —

H,O-— 0,44 DSS

P.O; 33,16 33,97

CO, 3,81 3,28

F 3,44 O57

SO; 0,17 0,44

ThA) a : 99,27 96,20

Analysts: Pelletal phosphorite—General Superintendence Co.

Hoedsjiespunt phosphorite— Anglo American Research Laboratories.

240 ANNALS OF THE SOUTH AFRICAN MUSEUM

ALUMINIUM PHOSPHATES

The aluminium phosphates will be discussed only briefly as they are not

directly related to the marine phosphorites, although both were related to the

upwelling of phosphorus-rich waters. These aluminium phosphates have been

described in some detail by Du Toit (1917) and Hutchinson (1950).

The aluminium phosphate distribution map (Fig. 25) shows how outcrops

of the phosphate rock occur sporadically about granite hills north and south of

Saldanha Bay. They are not confined to any particular altitude. The degree of

phosphatization is very variable over the area and even varies within any

particular hand specimen. On Konstabelkop the phosphate rock is hard and of

< Roof ( SALDANHA

(Ss

33;

GS 1S Hoedjiespunt

9 Bomgat

SALDANHA

Malgaskop

Sen oS

LIS

Lyfsershoek

<a ¢

kilometers C 4

Konstabelkop

O 5

Ko aS,

reefte Baay

Contour Interval 30m

33°10.

18°

Fig. 25

Distribution map for aluminium phosphate.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 241

moderate yellowish brown (1OYR 5/4) coloration, while the phosphatized clay

of slickensides is greyish brown (SYR 3/2). North of Saldanha Bay the phosphate

rock has a green hue. On Konstabelkop an exploration pit in a marine terrace at

152 m a.s.l. shows rounded boulders lying on weathered porphyry. Both the

underlying porphyry and the rounded boulders have been phosphatized, but in

the case of the boulders the degree of phosphatization is greatest near the surface,

suggesting post-depositional phosphatization.

In the main quarry on Konstabelkop the highest concentration of P,O, is

close to the surface, the concentration decreasing irregularly with depth. The

maximum depth ranges from 1,5 m to 12 m, the base being highly irregular. At

the surface weathering has produced a laterite-like layer.

The numerous slickensides, along which phosphatization is extensive, are

a feature of the Konstabelkop phosphate rock. In thin-section it is seen that the

groundmass is a fine clay material which has been phosphatized. The phos-

phatized parts of the clay are isotropic while slight anisotropism is evident at the

contact of the phosphatized and unphosphatized clay. Larger quartz grains are

generally fractured and have fissures filled with the same isotropic phosphate

mineral. The faulting appears to have followed the phosphate mineralization

since the texture of the rock shows the drag effect of movement along the fault

plane. The clay minerals must originally have developed along joint planes in the

granite. A detailed thin-section study was carried out and reported by Du Toit

(1917).

The altitude of the phosphate deposits is very variable (as shown in Fig. 25)

and cannot be related to any particular sea level as Visser & Schoch (1973)

maintain. On Baviaansberg within the grounds of the Naval Academy (SAS

SALDANHA) the phosphates occur at 45 m a.s.l. and have imparted a green

coloration to the granite. On Malgaskop the phosphates range from 40 to

100 m a.s.l. The lower limits of these phosphate deposits are not significant as

they have originated by percolating solutions.

TABLE 3

Partial chemical analyses of the aluminium phosphate rock.

Phosphatized Phosphatized Yellow sandy Phosphatized

porphyry porphyry regolith limestone

% % % Yo

DiOre ee es 59,20 65,40 10,72 10,70

NAO seis ee 8,32 6,78 31,11 5,66

PS (On | cee en 4,02 29 7,14 0,43

CAOP es 0,10 —- 0,42 40,90

ROOM es 0,60 0,48 2,41 0,60

PO) ee 12,14 11,68 18,31 0:92

COT a ec — — — 31,40

LE)” OS Trace Trace Trace Trace

Analyst: A.E. & C.I. Ltd. Results supplied by Mr Botha of the Konstabelkop Mine.

242 ANNALS OF THE SOUTH AFRICAN MUSEUM

Partial chemical analyses of the phosphate rock are listed in Table 3. Several

significant features are apparent:

(1) The Fe,O, content is high on the granite areas but insignificant on the

phosphatized limestone.

(2) Al,O, is highest in the yellow sandy regolith (31,11 per cent). Also in

this horizon the Fe,O, value is highest (7,14 per cent).

(3) Fluorine is always present in only trace quantities.

The yellow sandy regolith has resulted from lateritic alteration of the under-

lying phosphatized granite. Under normal soil-forming processes weak solutions

from the leaching of the rock evaporate and the least soluble components pre-

cipitate first. These include hydroxides of iron and aluminium, silica and

carbonates. Further solution leaves behind the aluminium and iron hydroxides

in an insoluble state.

GENESIS OF THE PHOSPHATE DEPOSITS

PHOSPHORITE

Authigenic phosphorites are the subject of a voluminous literature and many

authors have remarked on the close relationship between phosphorite deposits

and areas of active upwelling of nutrient-rich water (McKelvey et al. 1953;

McKelvey 1959, 1963; Sheldon 1964; Tooms et al. 1969). Since very few phos-

phorites are known to be forming at present, the origin of the phosphorite can

only be inferred from the geologically young deposits. Geologically young

phosphorite is commonly found in sediments adjacent to areas of modern

oceanic upwelling (McKelvey 1963). These are mainly on the west coasts of

continents but also to a limited degree on other coasts. Such areas of active

upwelling and phosphorite occurrence lie between the 40th parallels (Sheldon

1964), e.g. south-western Cape Province, Morocco, South America, California.

Ancient phosphorites on the other hand are found at higher latitudes, their

present distribution being the result of continental drift. Pevear (1966) has shown

that upwelling is certainly not a prerequisite for phosphate enrichment. He has

suggested an estuarine origin for the phosphate of the Phosphoria Formation of

the U.S.A. Such an estuarine environment would be an area of high biologic

productivity.

Brongersma-Sanders (1957) has related marine mass mortalities, and

phosphorite deposition, to regions of upwelling. Accumulation and decay of

organic matter in the oceans at depths below 500 m concentrate phosphorus, and

with upwelling caused by trade-winds blowing the surface water offshore, the

phosphorus-rich water is brought to the surface, where deposition of phosphorite

occurs. Precipitation of the phosphate occurs when the amount of phosphate

supplied exceeds the saturation value of the sea water (Bushinsky 1966). Precipi-

tation of the phosphorite from such saturated water is accelerated by increased

temperature and pH (Sheldon 1964). A rise in temperature drives off CO,, which

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 243

results in a rise in pH. A low rate of supply of terrigenous detritus is very impor-

tant (Pevear 1966). Tooms ef al. (1969) have found a close association between

ancient phosphorites and arid areas. Such areas would also be expected to have

low or negligible rates of sedimentation and are usually associated with coasts

with active upwelling. Bushinsky (1966) and D’Anglejan (1967) suggest that

phosphorites have formed in water shallower than 100 m, while Parker &

Simpson (1972) have found the greatest concentration of phosphate nodules on

the Agulhas Bank between 100 and 140 m.

The phosphorites of the Langebaanweg—Saldanha area have much in com-

mon with other phosphorite deposits. The phosphorite occurs authigenically in

the Miocene basal bed. Although this horizon has been truncated at 30 m above

sea-level it has been inferred that marine limestone at 56 m above sea-level north

of Saldanha represents the farthest inland extent of that sea (Tankard in press).

At Hoedjiespunt the phosphorite occurs at 5 m above sea-level, suggesting that

the depth of water at time of formation was about 50 m.

A common characteristic of phosphorite accumulations is their location on

one sice of a basin where deep phosphate-rich waters are upwelling adjacent to a

shallow shelf (Blatt et a/. 1972). Although uneconomic concentrations of phos-

phorite do occur on the open Atlantic coast between Saldanha Bay and St.

Helena Bay, the highest concentrations are found in the erosional basin between

Langebaanweg and Saldanha. The nearly ubiquitous presence of organic matter

within the pellets suggests high productivity and a common origin for the

organic matter and the phosphorite. At the base of the Varswater Formation on

Witteklip the pelletal phosphorite is black due to excessive quantities of carbona-

ceous matter (Fig. 6). The quantity of organic matter in the phosphorite decreases

upwards in the sedimentary sequence, and there is also a decrease in organic

matter in the phosphorite in the Langeberg direction, resulting in lighter-

coloured phosphorite. The trend is for a decrease in carbonaceous phosphorites

shoreward. Commenting on an identical situation in the Phosphoria Formation

(U.S.A.) Gulbrandsen (1969) notes: ‘Although the locus of major apatite pro-

duction is coincident with the locus of organic-matter accumulation, as exempli-

fied by the black carbonaceous phosphorites of the Phosphoria Formation

significant production can occur shoreward, where the organic matter is largely

destroyed, and form light-colored, low-organic-content phosphorites such

as indicated by the facies of the Phosphoria Formation. . . .. Summerhayes

(1973) notes that the only significant local concentration of phosphate on the

Agulhas Bank, other than that provided by the erosion of Tertiary phosphorite,

is in sediments enriched in non-skeletal organic matter.

If the phosphorite can be attributed to upwelling of cold, nutrient-rich

water, then it is also likely that the coastlands were as arid as at present. The

average annual rainfall for this area is about 260 mm, which occurs fairly

steadily in the winter months, and continental run-off is low. The relatively low

winter rainfall, low run-off, and the low-lying nature of the Sandveld would

promote low sedimentation rates. The intermittent reworking of the basal bed

244 ANNALS OF THE SOUTH AFRICAN MUSEUM

deposits and the high grades of phosphate suggest an environment of negligible

sedimentation.

In the Miocene basal bed on Langeberg the phosphorite occurs as an inter-

stitial component of a quartzose sandstone (packstone). Frequently the phos-

phorite occurs as drusy-encrustations about the quartz grains, indicating crystal

growth from the grain surfaces, the remaining voids are filled with clear collo-

phane and argillaceous and organic material. The structureless aspect of the

phosphorite pellets, the rare occurrence of nuclei and the mixing of collophane

with organic, argillaceous and ferruginous material suggest precipitation within

the interstices of the sediment. Ames (1959) found that the replacement of

calcium carbonate was possibly the only way in which the carbonate fluorapatites

could form. Rooney & Kerr (1967) have suggested that the North Carolina

phosphorite probably originated by replacement of calcareous matter and by

chemical precipitation under reducing conditions in a large, shallow lagoon or

estuary with a restricted circulation. Pevear (1966) suggests that the reason why

phosphorite is not forming today in Georgian estuaries is because virtually no

calcium carbonate is forming at present. Thin-section study of the Langeberg

basal bed rock has shown no evidence of a replacement mechanism, although

phosphatization of the calcite foraminiferal tests and mollusc shell fragments

demonstrate that such replacements have taken place in the overlying strata. At

Hoedjiespunt the apatite has only 3,60 per cent SiO,, 3,20 per cent of this pro-

bably being bound up with Al,O, as a clay material. Thin-section study confirms

that there is less than 1 per cent quartz. It is quite conceivable that the carbonate

apatite, in this case dahllite, originated by replacement of micrite and micro-

coquina. The negligible amount of quartz suggests extremely low sedimentation

rates in a sheltered environment. Tooms ef al. (1969) have shown that the syn-

thesis of carbonate apatite from solution means that the prior existence of

calcium carbonate is not a prerequisite to formation of this mineral. At Hoedjies-

punt the phosphorite has an oolitic structure in places while some structures

in the basal bed at Langeberg suggest that precipitation by algae has played

a small part. Charles (1953) has attributed much phosphorite formation to

precipitation by algae.

Summerhayes (1970) has noted the limited evidence for contemporaneous

formation of carbonate apatite, there being very few submarine phosphorites of

recent age. McKelvey et al. (1953) found no bedded phosphorites younger than

late Tertiary. The environmental conditions most favourable to phosphorite

formation occurred in the warmer mid Tertiary seas (Tooms ef a/. 1969). Baturin

(1971) has shown some evidence of phosphorite forming at present along the

South West African coast, these sediments being further discussed by Summer-

hayes et al. (1973).

Haughton (1932) attributed the phosphorites of Langebaanweg to alteration

of a calcareous deposit by phosphatic solutions. He thought the aluminium

phosphate and the phosphorite had a similar phosphate source, in that both

were derived from organic masses such as guano. Haughton saw three possible

PHOSPHORITE AND ALUMINIUM PHOPHATE ROCK 245

ways in which the phosphate could have arrived in the sediments at

Langebaanweg;

(1) the phosphate was derived from the guano deposits about Vredenburg

by percolating waters

(2) the area in question could be covered by continuously replenished

guano-dust blown in by west winds

(3) the guano was deposited on islands at Langebaanweg, but he notes

that no evidence of the guano can be found in the area.

If the phosphate was transported by groundwater from Vredenburg it must

be remembered that the Langeberg deposits are 13 km from that area. It has

already been shown that the phosphorite is essentially a basin deposit in the

study area, and that maximum concentration of phosphorite is about the

perimeter of the basin. Groundwater from Vredenburg would have had to

phosphatize selectively the periphery of the basin. Phosphorite deposits are also

found along the Namaqualand coast where topography is not conducive to

guano accumulation.

The phosphate is unlikely to have been transported as a fine guano-dust

since westerly winds blow only in the wet winter months in the south-western

Cape, and are rain-bearing. Dust is unlikely to be transported in large quantities

during this season. Furthermore, guano deposits do not lend themselves easily

to dust formation on the scale envisaged.

Certainly on the granite hills north and south of Saldanha Bay percolating

phosphate solutions have produced aluminium phosphates. The mineral pro-

duced has depended entirely upon the rock being phosphatized, e.g. on Kon-

stabelkop the phosphatization of the granite and limestone. On Witteklip—Sand-

heuwel the phosphorites are banked against the same granite masses but now the

phosphate is a carbonate fluorapatite just as at Langeberg. Similarly the phos-

phorite at Hoedjiespunt is a carbonate apatite, although it rests on a granite

shelf which it has phosphatized. It has been shown that the basal bed at Lange-

berg contains the phosphorite as an interstitial material and that there is no

evidence of calcium carbonate replacement. Chemically the pelletal phosphorite

of the Varswater Formation and the aluminium phosphates of the granite terrain

are totally different. The phosphorite characteristically has a higher F content,

while the aluminium phosphates contain higher concentrations of Al,O, and -

Fe,O3. If these deposits had a similar origin then the pelletal phosphorites

should include an aluminium-rich component. Frankel (1943) has noted the small

percentage of iron and aluminium phosphates. Frankel suggests that the sources

for the fluorine must be found in either the sea or the granitic areas. The latter

option must be discarded since the aluminium phosphates on the granitic rocks

are notably poor in fluorine. Frankel concludes that phosphatic solutions in

percolating through these deposits ‘converted the limestones and calcareous

nodules into (probably) hydroxy-apatite and the argillaceous rock types into

aluminium and iron phosphates.

246 ANNALS OF THE SOUTH AFRICAN MUSEUM

ALUMINIUM PHOSPHATE

The aluminium phosphate on the granite hills north and south of Saldanha

Bay is very different from the marine phosphorites of the Varswater Formation,

although it, too, can be related indirectly to upwelling phenomena. The upwell-

ing of nutrient-rich waters forms a food chain upon which great colonies of sea-

birds thrive. The largest guano deposits are found associated with such areas

(Hutchinson 1950). On the Cape west coast the most important guano birds are

Morus capensis (Cape gannet), Spheniscus demersus (jackass penguin) and Phala-

crocorax capensis (Cape comorant). Seabird guano is a richly nitrogenous

phosphorus material, the phosphate concentration increasing as the more

soluble nitrates are leached out. Flack (1916) reports the total nitrogen and

P.O, contents of mixed guano from Malagas, Marcus and Jutten Islands (Fig. 1)

as 10,94 per cent and 13,80 per cent respectively. CaO for this sample was 12,49

per cent. 4,11 per cent of the P.O; was water soluble, the rest being acid soluble.

Fresh guano is characteristically rich in nitrogenous phosphorus, but with

leaching of soluble nitrates the P.O; concentration increases. Low-pH waters

transport the phosphate from the guano to the bedrock where it reacts to form

new minerals. Thus reaction with limestone has produced calcium phosphate

and reaction with clay minerals formed along joint planes in the granite on

Konstabelkop has produced the aluminium phosphates. Whereas the phos-

phorites have high carbonate and fluorine contents, these are dependent entirely

upon the type of rock that is being phosphatized in the case of the aluminium

phosphates. Although the original source of the phosphate can be attributed to

guano accumulations, direct evidence of a guano deposit has been destroyed by

weathering. The formation of phosphorite, on the other hand, appears to be

largely due to precipitation from phosphate-rich sea water. At Hoedjiespunt the

phosphorite has resulted from post-depositional phosphate mineralization of

a micrite and microcoquina. While the precipitation of the phosphorite is

temperature and pH dependent, the aluminium phosphates appear to be depen-

dent upon large seabird colonies. Upwelling of nutrient-rich waters could at

present sustain vast seabird colonies but so far there is no evidence of phos-

phorite deposition. Thus, although the two phosphate deposits are dependent

upon the upwelling, there is no reason why the phosphorites and aluminium

phosphates should be related in time. It is more likely that the aluminium phos-

phates are derived from leaching of guano deposits which have accumulated

over a considerable length of time. Visser & Schoch (1973) believe that these

phosphates can be correlated with particular sea-levels, but if they owe their

origin to seabird colonies no such relationship seems likely.

According to Harrington ef al. (1966) aluminium phosphates result from

the solution and transportation of phosphate minerals from phosphatic lime-

stone or guano by low-pH waters in hot humid regions. Hutchinson (1950) and

Harrington et al. (1966) have shown that the largest guano deposits are found in

the vicinity of the same regions of upwelling of cold phosphorus-rich water as

the marine phosphorites. Harrington ef al. infer that where the bedrock consists

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 247

of silicates the phosphate will form a variety of aluminium silicate minerals, such

as variscite or metavariscite and crandallite. Du Toit (1917) found that the

average composition of the Konstabelkop phosphate is close to that of barrandite

while minute yellowish crystals may be referred to wavellite or variscite.

Altschuler et al. (1956) have found that the upper part of the Pliocene Bone

Valley Formation, Florida, has been altered to aluminium phosphate in a zone

averaging about 2 m in thickness. This alteration has taken place by weathering

and groundwater which has produced a progressive change in mineralogy with

depth. They found that the top of the zone is characterized by the aluminium

phosphate wavellite while the middle zone is characterized by the calcium

aluminium phosphates crandallite and millisite. Both crandallite and millisite

are virtually isotropic. The Bone Valley phosphates are characterized by higher

CaO and F than the Konstabelkop phosphates.

SUMMARY

Bedded marine phosphorites of late Tertiary age are found today in warm

climates between the 40th parallels in areas adjacent to divergent upwelling of

nutrient-rich waters. Precipitation of the phosphate is dependent upon an

increase in temperature as the upwelling water reaches the surface, and low rates

of supply of terriginous detritus. At Langeberg a consolidated phosphatic sand-

stone of Miocene age has formed by precipitation of the phosphate (francolite)

in the voids of a marine sand. The francolite is associated with finely divided

argillaceous, ferruginous, and carbonaceous material. On Hoedjiespunt, on the

other hand, the bedded apatite contains a total of only 3,6 per cent SiO,, most of

this being present in a clay mineral. The Hoedjiespunt apatite is defined by the

concentration of fluorine and carbonate as dahllite. At both Hoedjiespunt and

Langeberg sedimentation rates were very low in a sheltered environment. In the

phosphatic sandstone at Langeberg the francolite has grown from the quartz

grain surfaces, the voids being finally filled with isotropic collophane. The basal

bed also shows reworking of the sediment, and possibly precipitation by algae.

Deposition of the Miocene basal bed was followed by a period of emergence.

Then in the Pliocene the basal bed was partially reworked by a further trans-

gression, and the Varswater Formation deposited. Erosion of the basal bed

liberated the matrix material, which then became associated with the detrital

component of the Varswater Formation as pelletal phosphorite. The pelletal

phosphorite is very similar to the authigenic phosphorite of the basal bed in

texture. It contains much argillaceous, ferruginous and carbonaceous material as

well as silt-size quartz particles. Also included with the pellets are biogenic

remains, foraminiferal tests, minute fish teeth, etc. It was shown that the highest

organic carbon content is found at the base of the Pelletal Phosphorite Member

on the seaward side of the basin, and that organic carbon content decreases

shoreward.

248 ANNALS OF THE SOUTH AFRICAN MUSEUM

The phosphorite deposits are briefly compared with the aluminium phos-

phates which occur on the granite hills north and south of Saldanha Bay.

Whereas the phosphorite has formed as a marine precipitate, the aluminium

phosphates probably originated by leaching of guano deposits and subsequent

reaction of these solutions with the bedrock.

ACKNOWLEDGEMENTS

The writer wishes to express his sincere thanks to Chemfos Ltd. for permis-

sion to publish this paper, and to Mr G. Benfield for supplying borehole samples.

For discussions and advice on the manuscript I wish to thank Dr Q. B. Hendey

and Mr A. Ruddock. The assistance of Mr R. Cross with the scanning electron

microscope at Rhodes University is gratefully acknowledged. All photographs,

except the scanning electron photomicrographs, were printed by Mr N. Eden,

while Mr V. Branco assisted with the illustrations. The manuscript was very

kindly typed by Miss J. Harding. Chemical analyses were undertaken by Anglo

American Research Laboratories and the Geological Survey.

REFERENCES

ALTSCHULER, Z. S., CLARKE, R. S. & YOUNG, E. J. 1958. Geochemistry of uranium in apatite

and phosphorite. Prof. Pap. U.S. geol. Surv. 314—D: 45-90.

ALTSCHULER, Z. S., JAFFE, E. B. & Cuttita, F. 1956. The aluminium phosphate zone of the

Bone Valley Formation, Florida, and its uranium deposits. Prof. Pap. U.S. geol. Surv.

300: 495-504.

Ames, L. L. 1959. The genesis of carbonate apatites. Econ. Geol. 54: 829-841.

ARAKAWA, K. Y. 1971. Studies on the faecal pellets of marine invertebrates (excluding mol-

luscs). 1. Publs Seto mar. biol. Lab. 19: 231-241.

BATuRIN, G. N. 1971. Stages of phosphorite formation. Nature, Lond. (Phys. Sci.) 232: 61-62.

BLATT, H., MIDDLETON, G. & Murray, R. 1972. Origin of sedimentary rocks: 541-557. New

Jersey: Prentice-Hall.

BRONGERSMA-SANDERS, M. 1957. Mass mortality in the sea. Mem. geol. Soc. Am. 67: 941-1010.

BUSHINSKY, G. I. 1966. Old phosphorites of Asia and their genesis. Jerusalem: Israel Program

for Scientific Translations. (1968 translation.)

Cayeux, L. 1934. The phosphatic nodules of the Agulhas Bank. Ann. S. Afr. Mus. 31: 105-136.

CHARLES, G. 1953. Sur l’origine des gisements de phosphates de chaux sédimentaires. Jnt. geol.

Congr. 19 (11): 163-184.

D’ANGLEJAN, B. F. 1967. Origin of marine phosphorites off Baja California, Mexico. Mar.

Geol. 5: 15-44.

Du Tort, A. L. 1917. Report on the phosphates of Saldanha Bay. Mem. geol. Surv. S. Afr

10: 1-38.

FLAck, E. V. 1916. Some factors influencing the solubility of phosphoric oxide in mixed

fertilisers containing superphosphates. S. Afr. J. Sci. 13: 201-208.

FoLk, R. L. 1962. Spectral subdivision of limestone types. Mem. Am. Ass. Petrol. Geol.

1: 62-84.

FRANKEL, J. J. 1943. The constitution of the Langebaan phosphate rock. S. Afr. J. Sci. 39:

102-108.

GULBRANDSEN, R. A. 1969. Physical and chemical factors in the formation of marine apatite.

Econ. Geol. 64: 365-382.

HARRINGTON, J. F., WARD, D. E. & McKeELvey, V. E. 1966. Sources of fertilizer minerals in

South America: a preliminary study. Bull. U.S. geol. Surv. 1240: 1-66.

HAUGHTON, S. H. 1932. On the phosphate deposits near Langebaanweg, Cape Province.

Trans. geol. Soc. S. Afr. 35: 119-124.

PHOSPHORITE AND ALUMINIUM PHOSPHATE ROCK 249

HuTCHINSON, G. E. 1950. Survey of existing knowledge of biogeochemistry. 3. The biogeo-

chemistry of vertebrate excretion. Bull. Am. Mus. nat. Hist. 96: 1-554.

McConneLL, D. 1938. A structural investigation of the isomorphism of the apatite group.

Am. Miner. 23: 1-19.

McKeE ;vey, V. E. 1959. Relation of upwelling marine waters to phosphorite and oil. Bull. geol.

Soc. Am. 70: 1783-1784.

McKe .vey, V. E. 1963. Successful new techniques in prospecting for phosphate deposits. Jn:

SCIENCE, TECHNOLOGY & DEVELOPMENT; UN conference on applications of science and

technology for the benefit of the less developed areas. 2. Natural resources: 163-173.

Washington: US Agency for International Development.

McKE Lvey, V. E. 1967. Phosphate deposits. Bull. U.S. geol. Surv. 1252-D: 1-21.

McKELveY, V. E., SWANSON, R. W. & SHELDON, R. P. 1953. The Permian phosphorite deposits

of western United States. Int. geol. Congr. 19 (11): 45-64.

MurRRAY, J. & RENARD, A. F. 1891. Report on deep-sea deposits based on the specimens col-

lected during the voyage of HMS Challenger. Rep. Voy. Challenger 1873-1876 (Deep-sea

deposits): 391-400.

ParRKER, R. J. 1971. The petrography and major element geochemistry of phosphorite nodule

deposits on the Agulhas Bank, South Africa. Bull. SANCOR mar. geol. Progm. 2: 1-94.

PARKER, R. J. & SIESSER, W. G. 1972. Petrology and origin of some phosphorites from the

South African continental margin. J. sedim. Petrol. 42: 434-440.

PARKER, R. J. & Simpson, E. S. W. 1972. South African Agulhas Bank phosphorites. Phosph.

Potassium 58: 18, 27.

PevEAR, D. R. 1966. The estuarine formation of United States Atlantic coastal plain phos-

phorite. Econ. Geol. 61: 251-256.

Rooney, T. P. & KERR, P. F. 1967. Mineralogic nature and origin of phosphorite, Beaufort

County, North Carolina. Bull. geol. Soc. Am. 78: 731-748.

SHELDON, R. P. 1964. Paleolatitudinal and paleogeographic distribution of phosphorite. Prof.

Pap. U.S. geol. Surv. 501-C: 106-113.

SUMMERHAYES, C. P. 1970. Phosphate deposits on the northwest African continental shelf and

slope. (Unpublished Ph.D. thesis, University of London.)

SUMMERHAYES, C. P. 1973. Distribution, origin, and economic potential of phosphatic sedi-

ments from the Agulhas Bank, South Africa. Trans. geol. Soc. S. Afr. 76: 271-277.

SUMMERHAYES, C. P., BIRCH, G. F., ROGERS, J., DINGLE, R. V. 1973. Phosphate in sediments

off southwestern Africa. Nature, Lond. 243: 509-511.

TANKARD, A. J. 1974. Varswater Formation of the Langebaanweg-—Saldanha area, Cape

Province. Trans. geol. Soc. S. Afr. (In press.)

TANKARD, A. J. (Jn press.) Chemical composition of the phosphorites from the Langebaanweg-

Saldanha area, Cape Province. Trans. geol. Soc. S. Afr.

Tooms, J. S., SUMMERHAYES, C. P. & CRONAN, D. S. 1969. Geochemistry of marine phosphate

and manganese deposits. Oceanogr. mar. Biol. 7: 48-100.

VIsseR, H. N. & ScuHocu, A. E. 1973. The geology and mineral resources of the Saldanha Bay

area. Mem. geol. Surv. S. Afr. 63: i-iv, 1-150.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FIscHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FIsCHER, P.-H., DUVAL, M. & RarFry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Koun, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 1960b. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

Anthony J. Tankard

PETROLOGY AND ORIGIN OF THE PHOSPHORITE

AND ALUMINIUM PHOSPHATE ROCK OF THE

LANGEBAANWEG—SALDANHA AREA,

SOUTH-WESTERN CAPE PROVINCE

5 O7.66

VOLUME 65 PART 9 AUGUST 1974

ANNALS

Ko); THE SOUTH AFRICAN |

‘MUSEUM |

CAPE TOWN

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 _ Band

August 1974 Augustus

Part, 9 Deel

THE AMPHIPODA OF SOUTHERN AFRICA

PART 4

THE GAMMARIDEA AND CAPRELLIDEA OF

tae CAPE PROVINCE EAST OF CAPE AGULHAS

C. L. GRIFFITHS

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 21, 3, 5, 7-8), 3(1-2, 5, t.-p.i.), 51-2, 5, 7-9),

6(1, t.-p.i.), 7(1-3), 8, 9(1-2), 10(1),

11(1-2, 5, 7, t.-p.i.), 24(2), 27, 31(1-3), 33

Price of this part/Prys van hierdie deel

R9,20

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1974

ISBN 0 949940 48 8

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

THE AMPHIPODA OF SOUTHERN AFRICA

PART 4

THE GAMMARIDEA AND CAPRELLIDEA

OF THE CAPE PROVINCE EAST OF CAPE AGULHAS

C. L. GRIFFITHS

C.S.I.R. Oceanographic Research Unit, University of Cape Town

(With 18 figures)

[MS. accepted 15 October 1973]

CONTENTS

PAGE

Introduction : , : j : So Pu 251

The collecting stations

Collections from the marine environment . 252

Collections from estuaries . : ; hh e259.

Systematics . F ‘ ; : ; . 264

Gammaridea 4 : j : 4 Sd ee 1S)

Caprellidea . { j ; : : Hi. 330

Summary . : : ; : : 1333

Acknowledgement : ; ; ; eS SIS)

References . : : : : 2 5 aga

INTRODUCTION

This paper forms the fourth part of a series dealing with the gammaridean

and caprellid amphipod fauna of Africa south of 20°S. Parts one to three

(Griffiths 1973 and 1974a and b) have covered southern Mocambique, southern

South West Africa and Natal respectively, while the present section deals with

the Cape Province east of Cape Agulhas (i.e. from 20°E to 30°E).

A feature of this region is the considerable number of estuaries to be found

there, many of which have been investigated by ecological survey teams from

the University of Cape Town. However, most of these studies have been cursory

and really only Knysna Estuary has been adequately worked on, resulting in a

paper on its ecology by Day, Millard & Harrison (1952).

The marine sampling coverage for the eastern Cape, on the other hand,

has been relatively thorough, over 700 marine samples being represented in the

University of Cape Town collections (compared with 200 from Mocambique,

250 from South West Africa and 350 from Natal). An emphasis on the Cape is

equally evident in the works of Stebbing (1908a, 1910a) and K. H. Barnard

(1916, 1925, 1940, 1955) and this increased collecting effort has revealed a

comparably larger haul of species (173, compared with 65, 81 and 115 from

Mogambique, South West Africa and Natal respectively). At this stage it cannot

be definitely stated whether the Cape fauna is richer than that of the other areas,

251

Ann. S. Afr. Mus. 65 (9), 1974: 251-336, 18 figs.

25) ANNALS OF THE SOUTH AFRICAN MUSEUM

or whether the increased number of species is merely due to increased collecting

effort. However, the amphipod faunas of temperate areas have, in general, been

found to be richer than those of more tropical zones.

The dominant ocean current in the eastern Cape is the south-westerly

flowing Agulhas Current, the inner margin of which tends to follow the con-

tinental shelf. At its centre the current flows at a rate of about three knots, but

this decreases rapidly with depth. As the continental shelf widens towards the

west the warm Agulhas water is progressively forced further offshore and inshore

counter-currents and cold-water upwellings become progressively more impor-

tant. Thus, while surface temperatures at the core of the Agulhas Current are

usually in the range 20-25°C, thermoclines and upwellings mean that bottom

temperatures are considerably lower. Over the Agulhas Bank itself a flow of

upwelled water from south-east to north-west maintains the bottom tempera-

ture at 10-13°C. Slightly higher figures are obtained to the east of the bank, so

that at Still Bay temperatures at 50 m fall around 15°C, dropping to 12°C at

200 m. As one progresses east there is a further rise so that at Port Elizabeth

the temperature at 50 m averages about 17°C while north of East London a

comparable figure would be 18—20°C. It must be stressed that the inshore regime

is subject to considerable variation caused by periodic upwellings and counter-

current intensities. For example, during periods of strong westerly winds the

whole Agulhas Bank region becomes an Atlantic Ocean province, with a con-

sequent drop in surface temperature.

As the collections of the University of Cape Town from the southern and

eastern Cape Province fall into a large number of different sections these are

discussed briefly below under the categories of collections from the open sea,

and those from estuaries. The collecting locations are shown on Figure 1.

THE COLLECTING STATIONS

Collections from the marine environment

(a) Still Bay shelf transect (SST)

The samples in this series were collected off Still Bay on 20 to 22 June 1972

and form part of the material for an analysis of benthic distribution being

undertaken by Dr J. G. Field of the C.S.I.R. Oceanographic Research Unit,

University of Cape Town. The amphipod material derived from this collection

has proved particularly diverse and interesting. A total of 66 species was

recovered from the 82 samples, which were taken in the form of a transect from

5 to 200 m. These species included eight new to science, as well as a number of

new records and rare species. The majority of these new species and new records

was recovered from the 200 m station which appears to represent a habitat

type not previously adequately sampled, since large numbers of new species

from various other groups were also recovered here.

The transect was originally planned to sample an area of shelly sand, but

THE AMPHIPODA OF SOUTHERN AFRICA 253

Umzimvubu

Fig. 1. The eastern Cape Province showing collecting areas referred to in the text.

AGULHAS BANK

a band of green mud was found to occupy a region from 50 to 80 m. The sandy

area from 5 to 50 m was not heavily populated by amphipods, although Urothoe

pulchella, Mandibulophoxus stimpsoni, Hippomedon onconotus and a few other

Species were reasonably common. The muddy zone from 50 to 80 m was very

sparsely populated with just a few Ampelisca brevicornis and Perioculodes longi-

manus and an occasional concentration of Siphonoecetes dellavallei. In the

area deeper than 80 m a great diversity of species was to be found, the 200 m

254 ANNALS OF THE SOUTH AFRICAN MUSEUM

station being particularly rich. Large Ampelisca were a feature of the area,

notably A. fusca, but also A. chiltoni and A. brevicornis. Mandibulophoxus

stimpsoni and Leucothoe richiardi were also well represented while the unusual

Concholestes armatus n. sp. was discovered roaming the surface in old scaphopod

shells.

SST station data

Catalogue No. Date Location Depth (m) Substrate

SST 1-17 20/6/72 35°22'S8/22°31E 200 Rock and sand

SST 18-24 20/6/72 35°06’S/22°15’E 120 Shelly sand

SST 29-37 21/6/72 34°40’S/21°39’E 80 Shelly sand

SST 40-45 21/6/72 34°25’S/21°28’E 50 Green mud

SST 47-57 21/6/72 34°24’S/21°27’E 30 Shelly sand

SST 58-60 21/6/72 34°23’S/21°26’E 10 Shelly sand

SST 61-66 21/6/72 34°23’S/21°26’E 15 Shelly sand

SST 67-73 21/6/72 34°23’S/21°26’E 20 Shelly sand

SST 74-75 22/6/72 34°23’S/21°26’E 20 Shelly sand

SST 76 22/6/72 34°23’S/21°26’E 15 Shelly sand

SST 77 22/6/72 34°22’S/21°26’E 5) Shelly sand

SST 81 22/6/72 34°46’S/20°25’E 81 Green mud

(6) Algoa Bay dredge (LIZ)

This code has been allocated to a series of some 40 grab and dredge samples

collected in Algoa Bay during April 1954. Eleven of these samples revealed

amphipods, 24 species being represented. The only species at all common in

the area were Ampelisca chiltoni, Ampelisca diadema, Cheiriphotis megacheles

and the ubiquitous Paramoera capensis.

LIZ station data

Catalogue No. Date Location Depth (m) Substrate

eZ 5/4/54 33°55’S/25°3VE 8-10 Mud

IZ 3 5/4/54 33°56’S/25°40’E 17-18 Sand

LIZ 13 6/4/54 33°58’S/25°38’E 7-8 Sand

LIZ 17 7/4/54 33°58’S/25°40’E 14 Stones

LIZ 19 7/4/54 33°58’S/25°42’E ai) Sand and shells

LS, 925) 11/4/54 34°00’S/25°44’E 39 Sand and shells

LIZ 29 11/4/54 34°00’S/25°42’E 5-7 Rock

LIZ 31 6/4/54 33°57'S/25°38’E 9 Limestone and clay

IeIZ 32 6/4/54 33°58’S/25°39’E 9 Stones

LIZ 37 6/4/54 33°58’S/25°39’E 9 Stones

LIZ 40 11/4/54 34°00’S/25°42’E 7 Rock

The MB code is carried by a series of 88 dredge samples collected in Mossel

Bay during January 1956. The substrate of the bay is predominantly sand,

although there are considerable outcrops of rock. Twenty-four amphipod

species are recorded from the area. Samples taken from rock showed the most

frequently encountered species to be Ceradocus rubromaculatus and Gammaropsis

atlantica as well as Leucothoe spp. and Caprella spp. Sandy areas contained

THE AMPHIPODA OF SOUTHERN AFRICA 255

numerous Cheiriphotis megacheles, Lysianassa ceratina and Ampelisca spp.,

while Paramoera capensis, the most common species in the bay, was found in

both rocky and sandy areas.

MB station data

Catalogue No. Date Location Depth (m) Substrate

MB 4 12/1/56 34°09’S/22°07’E 10 Shell, rock

MB 5 12/1/56 34°08’S/22°08’E 21 Fine sand

MB 10 12/1/56 34°04’S/22°13’E 19 Rock

MB 13 12/1/56 34°04’S/22°13’E 19 Rock

MB 16 13/1/56 34°11’S/22°10’E 16 Sand and rock

MB 20 13/1/56 34°08’S/22°07’E 13 Sand, shells, rock

MB 21 13/1/56 34°10’S/22°08’E 8 Sand

MB 23 13/1/56 34°08’S/22°07’E 12'5 Rock

MB 28 13/1/56 34°11’S/22°09’E 19 Rock

MB 32 15/1/56 34°09’S/22°07’E 10 Shelly sand

MB 33 15/1/56 34°08’S/22°07’E 19 Sand

MB 34 15/1/56 34°08’S/22°09’E 31 Sand

MB 37 16/1/56 34°09’S/22°10’E 31 Sand

MB 38 16/1/56 34°10’S/22°07’E 8,5 Sand

MB 40 16/1/56 34°10’S/22°08’E 9 Rock

MB 41 16/1/56 34°10’S/22°08’E 9 Rock

MB 45 17/1/56 34°10’S/22°09’E 17 Sand

MB 46 17/1/56 34°11’S/22°10’E 26 Sand

MB 50 17/1/56 34°11’S/22°09’E 10 Rock

MB 54 17/1/56 34°10’S/22°09’E 14 Rock, sandy patches

MB 57 17/1/56 34°10’S/22°09’E 9 Rock

MB 58 18/1/56 34°04’S/22°13’E ES Rock

MB 59 18/1/56 34°04’S/22°13’E eS Rock

MB 61 18/1/56 34°04’S/22°14’E 17-20 Coarse sand, shell, rock

MB 66 18/1/56 34°04’S/22°13’E 26 Sand and rock

MB 69 19/1/56 34°08’S/22°07’E 113} 5) Sand, rocky

MB 70 19/1/56 34°08’S/22°07’E 18 Sand

MB 71 19/1/56 34°08’S/22°07’E 12 Sand

MB 73 19/1/56 34°09’S/22°07’E 12 Rock, sand, shell

MB 75 19/1/56 34°08’S/22°07’E 1555 Sand

MB 77 20/1/56 34°11’S/22°06’E 24 Rock and sand patches

MB 80 20/1/56 34°05’S/22°11’E 20,5 Fine sand and mud

MB 82 20/1/56 34°10’S/22°09’E — (Plankton haul)

MB 84 21/1/56 34°11’S/22°10’E 29 Rock

MB 86 17/1/56 34°11’S/22°09’E 10 Rock

MB 87 17/1/56 34°10’S/22°09’E 14 Rock, sandy patches.

(d) Trawler stations (TRA)

Material collected during excursions by members of the Zoology Depart-

ment of the University of Cape Town on commercial trawlers are grouped

under this code. Although there are 44 TRA stations in the area under considera-

tion only three contain amphipods. Five species have been identified from these

samples, none of them at all common.

TRA station data

Catalogue No. Date Location Depth (m) Substrate

TRA 54 28/11/52 34°40’S/21°35’E 75 Rock

TRA 55 28/11/52 34°40’S/21°35’E 15 Rock

TRA 58 26/11/52 34°28’S/21°45’E 70 Sand and stones

256 ANNALS OF THE SOUTH AFRICAN MUSEUM

(e) South coast dredge (SCD)

Grab and dredge samples collected between 20°E and 30°E along the

southern Cape coast, and which do not form part of the specific studies men-

tioned above, are allocated to the SCD catalogue. To date there are some 400

samples in the series and included in the recorded fauna are 96 species of

amphipod. Of these species Mandibulophoxus stimpsoni and Perioculodes

longimanus have been the most commonly found, while other frequently

occurring species have been Ampelisca brevicornis, Ampelisca palmata, Aora

typica, Cheiriphotis megacheles, Gammaropsis atlantica and Photis and Urothoe

spp.

SCD station data

Catalogue No. Date Location Depth (m) Substrate

SCD 3 18/4/58 34°31’S/24°40’E 102 Rock

SCD 10 19/4/58 34°15’S/25°05’E 11 Rock and shell

SCD 20 26/5/58 34°07’S/23°23’E 46 Rock

SCD 24 26/5/58 34°46’S/23°27'E 110 Rock

SCD 34 21/5/58 33°03’S/27°56’E Si Sand, shells

SCD 41 19/5/58 32°15’S/28°S7’E 47 Rock

SCD 55 20/8/58 34°01’S/25°45’E 46 Rock

SCD 59 19/8/58 33°37'S/26°56’E 46 —

SCD 60 16/8/58 33°02’S/27°56’E 46 Rock

SCD 62 15/8/58 32°17’S/28°54’E 46 Rock

SCD 64 14/8/58 31°37’S/29°36’E 36,5 Mud

SCD 74 16/7/59 32°33’S/28°38’E 55 Sand and mud

SCD 81 16/7/59 32°43’S/28°28’E 58 Stones, shells

SCD 83 17/7/59 27°54’S/33°03’E 51 Sand, shells

SCD 93 17/7/59 33°03’S/27°5S’E 27 Rock

SCD 94-96 20/7/59 34°21’S/25°41’E 110 Shell

SCD 99 21/7/59 34°33’S/24°O1’E 130 Rock

SCD 100 21/7/59 34°33’S/24°O1’E 130 Rock

SCD 102 21/7/59 34°33’S/24°O1’E 130 Rock

SCD 103 22/7/59 35°07’S/22°15’E 120 Sand

SCD 104-5 23/7/59 34°33’S/21°28’E 67 Sand, shells

SCD 106 23/7/59 34°35’S/21°10’E 67 Rock

SCD 110 23/7/59 34°35’S/21°11’E 15 Sand, stones

SCD 115 26/11/59 34°54’S/22°12’E 106 Sand, shells

SCD 118 14/2/60 34°24’S/21°46’E 18 Rock

SCD 120 14/2/60 34°33’S/21°52’E cil Sand

SCD 122 14/2/60 34°40’S/22°00’E 93 Sand

SCD 124 3/6/60 34°26’S/21°48’E 67 Mud

SCD 127-8 3/6/60 34°37’S/21°56’E 87 Sand

SCD 131 3/6/60 34°48’S/22°06’E 100 Sand

SCD 135 26/11/59 34°29’S/21°49’E 73 Mud

SCD 138 28/8/60 34°35’S/21°S56’E 77 Shells

SCD 141 28/8/60 34°46’S/22°05’E 93 Sand |

SCD 146 28/8/60 34°46’S/22°05’E 93 Sand

SCD 148 28/8/60 34°59’S/22°18’E 106 Sand

SCD 151 2/6/60 34°55’S/21°26’E 91 —

SCD 159 25/11/60 34°03’S/25°59’E 84 Rock

SCD 160 25/11/60 34°03’S/25°59’E 84 Rock

SCD 172 24/11/60 33°58’S/25°41’E 4-11 Rock

SCD 179 24/11/60 33°58’S/25°41’E 4-11 Rock

SCD 181 30/11/60 34°20’S/23°31’E 110 Sand I

SCD 184 25/11/60 34°23’S/26°01’E 137 Sand, shells i

Catalogue No.

SCD 185

SCD 188

SCD 189

SCD 192-3

SCD 194

SCD 198

SCD 199

SCD 202

SCD 204

SCD 208

SCD 211

SCD 216

SCD 219

SCD 222

SCD 225

SCD 227

SCD 228

SCD 230

SCD 232

SCD 235-7

SCD 244-5

SCD 248

SCD 249

SCD 253

SCD 257

SCD 262

SCD 267

SCD 269

SCD 273

SCD 276

SCD 278

SCD 280

SCD 282-3

SCD 285

SCD 286

SCD 287

SCD 288

SCD 295

SCD 300

SCD 302

SCD 304

SCD 308

SCD 310

SCD 311

SCD 312

SCD 315

SCD 319

SCD 321

SCD 324

SCD 326

SCD 328

SCD 329

SCD 332

SCD 338

SCD 339

SCD 342

SCD 343

THE AMPHIPODA OF SOUTHERN AFRICA

Date

25/11/60

30/11/60

29/11/60

30/11/60

29/11/60

30/11/60

25/11/60

24/11/60

25/11/60

29/11/60

25/11/60

30/11/60

29/11/60

5/12/60

29/11/60

4/12/60

30/11/60

29/11/60

30/11/60

16/7/61

14/7/61

16/7/61

19/7/61

14/7/61

16/7/61

11/2/62

6/2/62

11/2/62

6/2/62

8/2/62

9/2/62

10/2/62

11/2/62

Location

34°13’S/26°04’E

34°10’S/23°32’E

34°05’S/23°23’E

34°04’S/23°25’E

34°07’S/23°31E

34°10’S/23°32’E

34°05’S/23°23’E

34°51’S/23°41’E

34°23’S/26°01’E

33°58’S/25°42’E

34°03’S/25°58’E

34°02’S/23°28’E

34°13’S/26°04’E

34°20’S/23°31’E

34°07’S/23°31’E

35°43’S/20°31’E

34°04’S/23°25’E

36°28’S/21°11’E

34°51’S/23°41’E

34°02’S/23°28’E

34°04’S/23°25’E

34°48’S/23°39’E

33°07’S/28°01’E

33°53’S/25°42’E

33°48’S/25°47’E

33°02’S/27°56’E

34°23’S/25°54’E

33°53’S/25°42’E

33°02’S/27°56’E

33°09’S/28°05’E

34°04’S/23°23’E

33°01’S/27°55’E

34°04’S/23°23’E

33°04’S/27°57’E

33°09’S/28°02’E

33°39’S/27°15’E

34°00’S/25°53’E

34/00’S/25°53’E

33°59’S/25°51’E

33°58’S/25°47’E

34°15’S/25°50’E

34°27’S/25°57’E

34°27'S/25°57’E

34°43’S/25°40’E

34°04’S/23°23’E

34°03’S/23°23’E

34°02’S/23°27’E

34°39’S/23°41’E

Depth (m)

2s)

Substrate

Sand

Mud

Sand

Mud

Sand

Mud

Sand

Sand, shells

Rock

Sand

Rock

Sand

Mud

Sand

Rock

Mud

Rock

Sand

Rock

Sand, rock

Sand, shells

Sand

Sand, rock

Rock

Sand, shells

Sand

Sand, shells

Shells

Sand

Rock

Mud

Sand

Sand, rock

Sand

(Glass buoy)

Shell

Sand

Mud

Shell, sand

Sand, shells

258

Catalogue No. Date Location Depth (m) Substrate

SCD 345 12/2/62 34°16’S/22°17’E 73 Sand, mud

SCD 347 12/2/62 34°10’S/22°15’E 54 Mud

SCD 348 12/2/62 34°09’S/22°10’E 36 Sand

SCD 349 12/2/62 34°09’S/22°09’E 18 Sand

SCD 350 13/2/62 34°28’S/21°S0’E 73 Sand

SCD 352 16/4/62 34°25’S/25°56’E 210 Mud, shells

SCD 353 6/2/62 33°04’S/27°57’E 84 Sand, shells

SCD 356 6/11/62 36°01’S/19°45’E 300 Sand, mud

SCD 359 19/11/62 34°48’S/22°51’E 120 —

SCD 366 2/12/62 33°50'S/25°47'E 36 Sand, rock

SCD 368 2/12/62 33° 5019/25°47-8: 36 Sand

SCD 370 4/12/62 33°59’S/25°45’E 44 Sand

SCD 373 4/12/62 33°59’S/25°51’E 36-54 Sand, shells

SCD 374 4/12/62 38°59'S/25751-E 54 Sand, shells

SCD 376 4/12/62 33°53’S/25°49’E 44 Sand

SCD 379 5/12/62 33°53’S/25°48’E 44 Sand

SCD 381 5/12/62 33°53’S/25°48’E 44 Sand

SCD 383 5/12/62 33°52/9/25°38 2 fl Sand

SCD 384 5/12/62 33°552(5/255391 7 Sand

SCD 388 8/12/62 34°04’S/23°23’E 46 Rock

SCD 391 8/12/62 34°05’S/23°23’E 11 Sand

SCD 392 9/12/62 35°08’S/22°02’E 125 Sand, shells

ANNALS OF THE SOUTH AFRICAN MUSEUM

(f) Shore stations

The collections falling into this group are the earliest made by the University

of Cape Town Ecological Survey and were intended to reveal zonation of the

intertidal fauna around the South African coast. The various collections are

denoted by the following catalogue codes (most of the stations were visited

only once, but where material was collected on subsequent visits this is indicated

by doubling the code letter for second visits and tripling it in the case of third

Visits).

Catalogue code Location Map reference

J Port St. Johns 31°38’S/29°33’E

H The Haven 32°14’S/28°5S’E

Q Qolora 32°38’S/28°26’E

L, LL, LLL East London 33°02’S/27°54’E

xX Kleinmont 33°33’S/27°04’E

K Kowie 33°36’S/26°54’E

nye Richmond 33°44’S/26'35’E

E Port Elizabeth 33°58’S/25°38’E

ZL, Jeffreys Bay 34°05’S/24°S5’E

T Storms River 34°02’S/23°54’E

R, RR Robberg 34°05’S/23°22’E

KN, KKN_ Knysna 34°05’S/23°04’E

V, VV Mossel Bay shore 34°11’S/22°09’E |

S, SS Still Bay shore 34°23'S/21°26’E :

AR Arniston 34°41’S/20°14’E

AG Cape Agulhas 34°50’S/20°10’E

Amphipod records at these locations are purely of a presence-absence type,

since the original purpose of the collections was to give an indication of

distribution.

THE AMPHIPODA OF SOUTHERN AFRICA 259

Species most frequently encountered intertidally in this area are Ceradocus

rubromaculatus, Hyale grandicornis, Jassa falcata, Lysianassa ceratina and

Paramoera capensis, which were found at virtually all the collecting stations. A

number of species appear to be restricted to the intertidal zone and have yet

to be found sublittorally. These include Ampithoe africana, Elasmopus pecteni-

crus, Palinnotus natalensis, Temnophlias capensis and Hyale and Talorchestia spp.

Collections from estuaries

(a) Estuaries near Port St. Johns (STJ)

The largest of the rivers in this area is the Umzimbuvu, which enters the

sea at the town of Port St. Johns. At the time of sampling (1950) the estuary

was muddy but became sandy near the mouth. Heavy silting had restricted the

fauna but Grandidierella bonnieroides and G. chelata were abundant among stones

near the mouth while G. /ingorum was dominant further upstream. Melita

zeylanica and Orchestia rectipalma were also to be found on the mud flats,

while Talorchestia and Orchestia spp. occupied the driftline. Urothoe pulchella

was also recovered, but was restricted to clean sand near the mouth.

Just south of the town of Port St. Johns lie two minor estuaries known as

the Eastern and Western Estuaries. Here again muddy zones were occupied by

such typical estuarine species as Melita zeylanica, Corophium triaenonyx and

Grandidierella spp. while Urothoe pulchella was to be found in clean sand.

The Umgazi Estuary, a few kilometres further south, was also sampled.

Here again the fauna conformed to the pattern typical of the area with Melita

zeylanica, Corophium triaenonyx and Grandidierella spp. giving way to Afrochil-

tonia capensis further upstream.

A brief exploratory visit to the Umgazana River 6 km south of the Umgazi

revealed a population of Orchestia rectipalma living amongst algae on the

pneumatophores of mangroves.

STJ station data

Catalogue No. Date Location

SEI 5 17/1/50 Drift line, Umzimvubu R. mouth

STJ 6 17/1/50 Intertidal mud, Umzimvubu R. mouth

STJ 7 17/1/50 Rocks, Umzimvubu R. mouth

STJ 8 17/1/50 Mud sievings, Umzimvubu R. mouth

STJ 14 18/1/50 Clean sand, Umzimvubu R. mouth

ST 15 18/1/50 Clean sand, Umzimvubu R. mouth

STJ 16 18/1/50 Western Estuary

SH 17, 18/1/50 Mouth of Eastern Estuary

STJ 18 18/1/50 Mud bank, 2 km up Umzimvyubu R.

STJ 24 19/1/50 Netting, 2 km up Umgazi R.

STJ 26 19/1/50 Muddy sand, 2 km up Umgazi R.

STJ 27 19/1/50 Under stones, Umgazi R. mouth

STJ 28 19/1/50 Mangrove roots, Umgazana R.

STJ 29 19/1/50 Sandy beach, Umgazi R. mouth

STJ 31 20/1/50 Decaying wood, Western Estuary

STJ 32 20/1/50 Weeds and stones, Western Estuary

260 ANNALS OF THE SOUTH AFRICAN MUSEUM

(b) Estuaries near The Haven hotel (HAV)

Three estuaries in the vicinity of The Haven were briefly visited by a party

of biologists from the University of Cape Town in January 1950.

The Bashee River winds across a narrow flood plain bordered by steep

wooded hills and then widens into a lagoon before entering the sea between

sand dunes. The area near the mouth is predominantly sandy but there are

occasional rocky outcrops, further upstream the sand is replaced by mud and

there is a dense growth of reeds and patches of mangroves. The muddy zones

were dominated by Grandidierella lignorum while the clean sand near the mouth

was occupied by Urothoe pulchella. Orchestia ancheidos occurred along the

drift line.

Blind Lagoon is a small, almost permanently closed estuary just below the

hotel. At the seaward end the bottom is sand but this is replaced upstream by

mud which forms the bottom of the rest of the lagoon except the head, where

there are outcrops of tock. The body of the estuary was well colonized by seven

species of amphipod, all typical estuarine forms. Most common of these was

Grandidierella chelata which was originally described from specimens collected

here.

The Mbanyana River Estuary consists of a large lagoon surrounded by

steep wooded banks and flowing into the sea through a narrow sandy channel.

At the time of sampling the depth of the lagoon was less than 30 cm and the

bottom graduated from sand at the seaward end through mud to stones at the

head. Near the outlet Corophium triaenonyx, Grandidierella lignorum and

Urothoe pulchella were common while further upstream they were replaced by

Melita zeylanica and Orchestia rectipalma.

HAV station data

Catalogue No. Date Location

HAV 3 12/1/50 Outlet of Mbanyana R.

HAV 5 13/1/50 Sand, Blind Lagoon

HAV 7 13/1/50 D-netting, Blind Lagoon

HAV 8 13/1/50 Submerged log, Blind Lagoon

HAV 9 13/1/50 Rocks, Blind Lagoon

HAV 10 13/1/50 Stones, Blind Lagoon

HAV 13 13/1/50 D-netting, mouth of Bashee R.

HAV 17 14/1/50 Sand, mouth of Mbanyana R.

HAV 18 14/1/50 Stones, 1 km up Mbanyana R.

HAV 20 14/1/50 Netting, 4 km up Bashee R.

The Keiskama River is tidal for some 30 km, widening over this distance

to some 2 km at Hamburg before entering the sea between headlands at

33°18’S/27°29’E. The river carries a good deal of silt and is shallow and muddy

except at the actual mouth where there are rocks and a small area of clean sand.

Only five species of amphipod have been recovered from the system. Of these

Orchestia ancheidos and Talorchestia capensis were locally common under

THE AMPHIPODA OF SOUTHERN AFRICA 261

weeds along the driftline, while Afrochiltonia capensis, Melita zeylanica and

Orchestia rectipalma were to be found under stones and among weeds throughout

the tidal reaches.

HAM station data

Catalogue No. Date Location

HAM 3 9/1/50 Zostera bed, 1 km from mouth

HAM 4 9/1/50 Driftline, 1 km from mouth

HAM 9 10/1/50 Under stones, Hamburg jetty

HAM 11 10/1/50 Hand netting, 8 km from mouth

HAM 13 11/1/50 Hand netting, 30 km from mouth

(d) Bushmans River (BMR)

A brief examination of the Bushmans River Estuary by a team from the

Zoology Department of the University of Cape Town in September 1950

revealed five amphipod species. The upper reaches were colonized by a typical

fauna of Grandidierella lignorum, Corophium triaenonyx and Melita zeylanica

with Orchestia ancheidos occurring along the banks. The only unusual record

was one of Ampelisca spinimana from Zostera near the mouth (33°41’S/26°41’E).

BMR station data

Catalogue No. Date Location

BMR 7 9/9/50 Digging and netting 30 km from mouth

BMR 21 12/9/50 Bank, 24 km from mouth

BMR 23 14/9/50 Zostera bed, 4 km from mouth

BMR 25 14/9/50 Reed bed, 40 km from mouth

BMR 26 15/9/50 Zostera bed, 4 km from mouth

(e) Sundays River Estuary (SUN)

The Sundays River flows between vertical mudbanks until about 1 km

from the wide shallow mouth (at 33°43’S/25°51’E) where the bottom becomes

sandy and the east bank rocky. A survey of the fauna revealed only three

amphipod species. Of these Urothoe pulchella and Corophium triaenonyx were

recovered from clean sand near the mouth, while Melita zeylanica occurred

amongst weeds on the piles of a bridge 8 km upstream.

SUN station data

Catalogue No. Date Location

SUN 5 7/1/50 D-netting, sand flats at mouth

SUN 6 7/1/50 General collection, 8 km from mouth

(f) Knysna Estuary (KNY)

The Knysna Estuary is the richest in southern Africa and has been the

subject of a detailed report by Day, Millard & Harrison (1952), who describe

the topography and fauna of the system, based upon expeditions to the area

between 1947 and 1955. These records have since been supplemented by a

further collection taken in 1964.

262 ANNALS OF THE SOUTH AFRICAN MUSEUM

The Knysna River lies 80 km east of Mossel Bay and widens into a large

tidal basin over the last 18 km before entering the sea between two massive

headlands at 34°05’S/23°04’E. The estuary is S-shaped and can be regarded as

beginning at Charlesford Rapids. Below the rapids lies a stony ford known as

‘The Old Drift’. From this point the river winds between marshy banks and

beneath the Westford road bridge and then widens abruptly. Around the banks

of the muddy upper basin lie the areas of Eastford, Ashford, Belvedere and

The Point, while the small Salt River enters the north bank just above the

railway bridge. Below the bridge the estuary is over 3 km wide and the channel

is fringed by extensive muddy banks covered in Zostera. There are two islands

in the lower basin, the upper of these, Paarden Island, lies just off Knysna town

and on it is situated Thesen’s wharf. The lower island, Leisure Island, is

connected to the mainland by a causeway crossing boggy salt marshes. Below

Leisure Island the lagoon narrows, the channel rapidly deepens and the banks

become rocky before entering the heads at Fountain Point.

The flow in the estuary is generally not strong but is sufficient to maintain

a salinity gradient. Tidal range at the heads is about 2 m, this range is maintained

as far as Westford bridge but falls to 0,3 m at Charlesford. Within the lagoon

wave action is negligible despite the vicious waves pounding the heads. Salinity

only begins to fall significantly above the rail bridge at which point values of

about 30%, are usual (although there may be significant layering effects). At

Westford bridge salinity ranges from 15-25%, while comparable figures at

Charlesford are 3-7%,.

Twenty-seven species of amphipod have been found in Knysna Estuary.

Most of these are not typical estuarine forms and are restricted to the area

below the railway bridge where salinity remains high. Most widespread in this

area are Cymadusa filosa, which lives in mucous tubes on Zostera plants, and

Paramoera capensis, Jassa falcata, Corophium triaenonyx and Lembos hypacan-

thus, which are locally abundant, especially on hard surfaces. Above the railway

bridge the amphipod fauna is dominated by estuarine species such as Grandi-

dierella lignorum, Melita zeylanica, Orchestia rectipalma and Corophium triae-

nonyx which all extend as far as Charlesford Rapids.

KNY station data

Catalogue No. Date Location

KNY 6 15/4/47 Dredge below Paarden Island

KNY 11 16/7/47 Dredge below Paarden Island

KNY 13 15/7/47 Channel west of Leisure Isle

KNY 28 17/7/47 Channel, Leisure Isle

KNY 30 16/7/47 Channel off Thesen’s wharf

KNY 42 18/7/47 Zostera bed, Salt River

KNY 43 17/7/47 Zostera bed, Leisure Isle

KNY 50 19/7/47 Westford bridge

KNY 57 20/7/47 Channel at Fountain Point

KNY 81 27/11/47 Belvedere

KNY 101 12/4/49 Below Charlesford

KNY 103 —/4/49 Westford bridge

THE AMPHIPODA OF SOUTHERN AFRICA 263

Catalogue No. Date Location

KNY 112 14/4/49 Below Charlesford

KNY 113 14/4/49 Zostera bed, ‘The Point’

KNY 114 14/4/49 Zostera bed, Leisure Isle

KNY 122 12/4/49 Westford bridge

KNY 128 14/4/49 Causeway below Woodbourne

KNY 139 12/4/49 Zostera bed, Paarden Island

KINY 157 27/3/50 Seaward side of Leisure Isle

KNY 160 —/11/47 Charlesford Rapids

KNY 162 9/7/50 Rail bridge

KNY 166 9/7/50 Leisure Isle and buoys in channel

KNY 171 9/7/50 Knysna Heads and Fountain Point

KNY 175 10/7/50 Charlesford Rapids

KNY 176 11/7/50 Buoy off Leisure Isle

KNY 179 11/7/50 Sandbank at Brenton

KNY 181 12/7/50 Old drift

KNY 184 12/7/50 Charlesford

KNY 187 13/7/50 Sandbanks, Leisure Isle

KNY 191 14/7/50 Fountain Point

KNY 245 14/2/64 Leisure Isle

KNY 266 14/2/64 Ashford

KNY 272 15/2/64 Rail bridge

KNY 273 15/2/64 Rail bridge

KNY 274 15/2/64 Rail bridge

KNY 283 15/2/64 Rail bridge

KNY 285 15/2/64 Old Drift

KNY 286 15/2/64 Old Drift

KNY 291 15/2/64 Old Drift

KNY 294 14/2/64 Ashford

(g) Estuaries near the Great Brak River (GBR)

Three estuaries in the vicinity of the Great Brak River (34°03’S/22°14’E)

were visited by the University of Cape Town Zoological Survey team in 1950.

The amphipod faunas of the three rivers conform to the pattern typical of the

area with Urothoe pulchella being found in clean sand near the mouth, while

Corophium triaenonyx, Melita zeylanica and Grandidierella lignorum occurred

throughout the estuaries. Talorchestia australis was found along the drift line.

GBR station data

Catalogue No. Date Location

GBR 12 30/4/50 Sand at mouth of Great Brak R.

GBR 13 30/4/50 Rocky shore near mouth, Great Brak R.

GBR 16 1/5/50 Sandy bottom of Great Brak R. 1 km from mouth

GBR 23 2/5/50 Above road bridge, Little Brak R. :

GBR 24 2/5/50 Head of estuary, Little Brak R.

GBR 37 3/5/50 Weeds at mouth, Great Brak R.

GBR 46 4/5/50 Mouth of Wilderness Estuary

(A) Breede River Estuary (BRE)

The Breede River is one of the largest in the Cape Province and has been

the subject of two expeditions by the University of Cape Town Ecological

Survey. Over the last 50 km of its course the river is tidal and the channel

reaches 8 mi in depth. The mouth (at 34°25’S/20°53’E) is permanently open with

264 ANNALS OF THE SOUTH AFRICAN MUSEUM

a sand spit and rocky areas to the north, while the southern shore displays

extensive mud and sand flats. At Karools Kraal, situated 7 km from the mouth,

the river is narrow and deep and runs between rocky banks interspersed with

muddy bays in which there are Zostera beds. Below this point the river flows

past Dolla se Baai and Green Point to Moddergat where the Zostera becomes

more extensive. Port Beaufort lies on the north bank about 2 km from the mouth

and Low Tide estate is opposite it on the south bank.

From this point to Witsands at the mouth the banks are mostly sand which

is locally muddy and covered with a luxuriant growth of Zostera at the lower

levels. At the mouth itself the south bank consists of wave-washed rocks which

display a true marine fauna.

Six species of amphipod have been found in the estuary. Of these Lysianassa

ceratina occurred only at the mouth and not under estuarine conditions. Melita

zeylanica and Orchestia rectipalma were found throughout the system while

Grandidierella lignorum was common amongst Zostera and Talorchestia capensis

along the drift line. Paramoera capensis was found in the Low Tide estate area.

BRE station data

Catalogue No. Date Location

BRE 5 1/7/51 Drift line, Port Beaufort

BRE 13 2/7/51 Zostera bed, Moddergat

BRE 18 2/7/51 Drift line, Moddergat

BRE 30 Si 7/5l Between Witsands and Port Beaufort

BRE 31 3/7/51 Between Witsands and Port Beaufort

BRE 34 4/7/51 Intertidal rocks, Port Beaufort

BRE 43 4/7/51 Rocks, Karools Kraal

BRE 44 5/7/51 Rocks at mouth

BRE 51 6/7/51 Zostera bed, Green Point

BRE 52 6/7/51 Zostera bed, Moddergat

BRE 55 6/7/51 Karools Kraal

BRE 56 6/7/51 Karools Kraal

BRE 57 6/7/51 Karools Kraal

BRE 71 8/7/51 Zostera bed, Green Point

BRE 77 8/7/51 Zostera bed, Dolla se Baai

BRE 81 8/7/51 From stomach of Lithognathus

BRE 123 5/2/52 Karools Kraal

BRE 128 5/2/52 Under stones, Karools Kraal

BRE 135 6/2/52 Among weeds, Low Tide estate

BRE 144 UPAlsy Rocks at mouth

BRE 146 8/2/52 Karools Kraal

SYSTEMATICS

Taxonomy of the Gammaridea followed here is that adopted by J. L.

Barnard (19695) and amended by J. L. Barnard (1970, 19725), while that of the

Caprellidea is modelled on the system proposed by McCain (1970). The arrange-

ment of families, genera within each family and then of species within each

genus is alphabetic. Limbs of the pereon are referred to as gnathopods | and 2

followed by pereiopods 1-5 (while most authors use this system some number

the pereiopods according to the segments on which they occur, i.e. gnathopods

1 and 2 followed by pereiopods 3-7). The analysis presented here is restricted

;

THE AMPHIPODA OF SOUTHERN AFRICA 265

to species occurring between the drift line and 1 000 m depth, thus estuarine

and beach-living species are included, while terrestrial and freshwater forms

are excluded, as are those found only at abyssal depths. Holotypes of all new

species have been placed in the South African Museum, Cape Town, paratypes

have been retained by the University of Cape Town.

For each species at least one reference has been given to what is considered

an accurate and, if possible, well-illustrated description. Full synonymies and

reference lists may be found through these descriptions. Where diagnoses are

provided they are intended to distinguish the species from others in the genus.

Generic and familial diagnoses are to be found in J. L. Barnard (19695), or may

be located through McCain & Steinberg (1970), in the case of the Caprellidea.

The sample coding system used here is that employed by the University

of Cape Town. Each area has its own catalogue indicated by a code of one,

two or usually three letters (MB = Mossel Bay, KNY = Knysna, etc.). Each

sample from that area is numbered and then each species within the sample is

denoted by a letter of the alphabet. Thus each specimen bears a catalogue/-

sample/species code. For example a sample from Mossel Bay is allocated to

the MB catalogue. The first sample in this catalogue is MB | and the species

from that sample called MB 1A, MB 1B, MB IC, etc. Where the number of

individual specimens is recorded this is indicated by a figure following the code

in brackets. Thus MB 1D(6) indicates that six specimens of the species D were

collected from station number | in the Mossel Bay series.

Authors working in this area in the past frequently gave collecting locations

in a somewhat vague manner, e.g. ‘Off Cape Agulhas’ or ‘5 miles SE of Cape

Infanta’. In presenting these records I have given the latitude/longitude square

in which they were made, followed by the depth and the reference from which

the records were taken. Thus 33°S/28°E/47m (Stebbing 1917) indicates that

Stebbing (1917) records the species in question from the 33°S/28°E area at a

depth of 47 m. In some cases material reported on by K. H. Barnard (1951,

1955, 1957) was derived from University of Cape Town collections and in these

cases only the University code is given.

Suborder GAMMARIDEA

Family Acanthonotozomatidae

Cypsiphimidia gibba K. H. Barnard, 1955

Cypsiphimidia gibba K. H. Barnard, 1955: 88, fig. 43.

Records: LIZ 40G(1).

Diagnosis: Pereon segment | swollen, its front margin nearly horizontal such

that the head projects vertically downwards; coxa | partially concealed by 2;

body entirely smooth; gnathopod | minutely chelate; gnathopod 2 subchelate;

telson short, apically incised.

Distribution: The above record is the only one to date.

266 ANNALS OF THE SOUTH AFRICAN MUSEUM

Dikwa n. gen.

Diagnosis: Upper lip slightly emarginate; mandible short and broad, with an

acute apex, spine row and large molar; lobes of lower lip not incised; palp of

maxilla 1 bi-articulate, exceeding outer plate; maxillipedal palp 4-articulate,

exceeding outer plate, article 2 of palp not produced; gnathopod 1 chelate;

gnathopod 2 simple; telson emarginate.

Type species: Dikwa acrania Nn. sp.

Relationships: The combination of chelate gnathopod 1 and simple gnathopod

2, together with the unusual mandible, demands the erection of this new genus.

Dikwa acrania n. sp.

Fig. 2

Description of female (3 mm): Head greatly reduced, shorter than first pereon

sey.nent and consisting largely of a rounded downturned rostrum (Fig. 2A),

eyes absent; upper lip emarginate; mandible (Fig. 2C) short and broad with a

3-articulate palp, articles 1 and 2 of palp subequal, their surfaces markedly

ridged, article 3 of palp the longest, a row of 14 strong setae along its medial

margin, incisor consisting of a sharp projection, lacinia mobilis apically bifurcate,

spine row of 13 strong spines, molar fairly large; lower lip not incised; maxilla

1 with bi-articulate palp exceeding outer plate, its distal margin setose, outer

plate terminating in eight strong serrate spines, inner plate bearing three terminal

setae; maxilliped with 4-articulate palp exceeding outer plate, none of the

articles distally produced, outer plate bearing a distal row of eight plumose

setae and a few simple setae, inner plate bearing a marginal and three sub-

marginal rows of minute pectinations; antennae subequal, slightly shorter than

pereon; article 1 of antenna | twice as wide as article 2 and as long as 2 and 3

together, apically lobed to partially envelop article 2, flagellum 14-articulate,

accessory flagellum absent; flagellum of antenna 2 15-articulate.

Coxae complex (Fig. 2A) but generally very thick and acuminate, coxa 7

very elongate and distally produced into an acute backwardly curved tooth;

pereon dorsally carinate, the carinae on pereon segments 6 and 7 produced

into teeth; gnathopod 1 (Fig. 2G) slender, chelate, article 2 with a small

protuberance on anterior margin, article 3 elongate, 6 as long as 3-5 together;

gnathopod 2 (Fig. 2H, I) very elongate and slender, simple, article 3 elongate,

7 tapering to an acute point which bears hooked setae; pereiopods | and 2

powerful, article 4 slightly produced antero-distally; pereiopods 3 and 4 with

article 2 greatly lobed postero-distally so as to obscure most of articles 3 and 4;

(pereiopod 5 missing on both sides).

Pleon segments strongly carinate mid-dorsally, each with a pair of latero-

dorsal humps; first pleonal epimeron postero-distally produced into a rounded

lobe, the second acutely produced, a rounded lobe in centre of posterior margin;

third pleonal epimeron postero-distally rounded, slightly produced; urosome

THE AMPHIPODA OF SOUTHERN AFRICA 267

Fig. 2. Dikwa acrania n. gen., n. sp.

Female, 3 mm: A—lateral aspect; B—upper lip; C—mandible; D—lower lip;

E—maxilla 1; F—maxilliped; G—gnathopod 1; H—gnathopod 2; I—tip of

article 7 of gnathopod 2 enlarged.

268 ANNALS OF THE SOUTH AFRICAN MUSEUM

strongly deflexed beneath pleon, segment | longer than 2 plus 3 and bearing a

medio-dorsal and a pair of latero-dorsal humps, segments 2 and 3 smooth;

uropods elongate, unarmed, rami narrow-lanceolate, projecting equally; telson

slightly longer than broad, apically emarginate.

The entire integument appears to be composed of small plates, in places

bearing a resemblance to the scales of a fish.

Colour: Uniform white (as preserved in 70°%% alcohol).

Holotype: SAM A13213, female, 3 mm.

Type locality: SST 11W, 32°22'S/22°31'E, 20 June 1972, depth 200 m, substrate

coarse khaki sand.

Remarks: This unusual species can easily be recognized by the markedly reduced

head. In life the antennae are flexed beneath the pereon so that the animal appears

to have had its head broken off.

Material; 2 9° from the type locality.

Iphimedia capicola K. H. Barnard, 1932

Iphimedia capicola K. H. Barnard, 1932; 118, fig. 66.

Records: SCD 160B(1), SCD 181T (1), SCD 216N(1); SST 16M(4).

Diagnosis: Rostrum acute, downturned; article 1 of antenna | terminating in

one dorsal and two ventral teeth, flagellum 12-articulate; pereon segments 1

and 7 much longer than other segments; pereon segment 7 and pleon segments

1-3 each with a pair of procumbent dorsal teeth; pleon segments dorsally keeled ;

postero-distal corner of third pleonal epimeron produced into an upturned

tooth, a second upturned tooth on posterior margin of the epimeron; telson

apically truncate, a pair of small denticles on each margin near the apex.

Distribution: Endemic to south and west coasts of South Africa.

Family Ampeliscidae

Ampelisca acris n. sp.

Fig. 3

Ampelisca excavata: K. H. Barnard, 1955: 82, fig. 40A.

(non) Ampelisca excavata K. H. Barnard, 1925: 336, pl. 34, figs 5-7. Gray & J. L. Barnard,

1970: 67-83, figs 1-5, pl. 1.

Description of female (11 mm): Head as long as two pereon segments (Fig. 3A),

antero-ventral margin oblique; two pairs of eyes with corneal lenses, the lower

pair just behind lateral angles of head, a small pigment spot behind upper eye;

antenna | extending well beyond peduncle of antenna 2, flagellum 19-articulate;

THE AMPHIPODA OF SOUTHERN AFRICA 269

Fig. 3. Ampelisca acris n. sp.

Female, 11 mm: A—head; B—pereiopod 3; C—pereiopod 5; D—lateral view of

urosome; E—uropod 3; F—telson.

270 ANNALS OF THE SOUTH AFRICAN MUSEUM

antenna 2 about half length of body, flagellum 20-articulate; mandibular palp

borne on a large process, article 1 half as long as 2, article 3 60% length of 2,

spine row of nine spines; outer plate of maxilla 1 with 11 spines, palp terminating

in four cusp teeth, four blunt spines and about 10 setae; inner plate of maxilliped

extending to tip of article 1 of palp, armed with two blade-like spines, outer

plate bearing 10 spines ranging from short and blade-like proximally to long

and evenly tapering distally.

Gnathopod | moderately setose, article 6 ovate, half as wide as long; dactyl

less than half length of article 6 and bearing three accessory setae; gnathopod

2 more slender than 1, article 6 half length of 5, dactyl less than half article 6

and bearing six accessory setae; pereiopod | not heavily setose, article 4 with

antero-distal angle slightly produced, article 7 equal to 5 plus 6; coxae 1-3 each

with a postero-distal tooth; article 2 of pereiopod 3 (Fig. 3B) as wide as long,

article 5 bearing four rows of spines posteriorly, the spines in the terminal row

serrate, article 6 with a row of four spines along its hind margin, dactyl bifurcate;

pereiopod 4 very like 3 but with a shorter article 6 which bears spines on its

anterior as well as posterior margin; pereiopod 5 (Fig, 3C) with an unusual

article 2 projecting to an acute point in line with the distal end of article 5,

posterior margin slightly concave distally, anterior edge of the lobe partially

obscuring articles 3 and 4, article 4 almost twice length of 3, its postero-distal

corner slightly produced and bearing four plumose setae, article 5 equal to 4

and bearing an antero-distal spine and a postero-distal group of three spines

and two plumose setae, article 6 slightly shorter than 5, two groups of spines

on outer margin, article 7 short and wide.

Third pleonal epimeron very slightly produced postero-distally; pleon

segment 4 weakly crested dorsally (Fig. 3D); uropod 1 projecting to end of

uropod 2, rami equal, subequal to peduncle, inner ramus dorsally spinose; rami

of uropod 2 equal, the outer dorsally spined; uropod 3 (Fig. 3E) with equal

rami, the outer slightly the more slender, apically acute and setose on both

margins, inner ramus with its upper margin cut into eight strong cusps, a

minute accessory tooth in each hollow, tip of ramus serrate, its lower-distal

margin bearing four plumose setae; telson 70% cleft (Fig. 3F), each lobe

bearing 2 long apical setae and with two pairs of smaller setae on the dorsal

surface.

Holotype: SAM A13206, female, 11 mm.

Type locality: MB 50Q, 34°11'S/22°09’E, 17 January 1956, depth 10 m, substrate

rocky.

Relationships: There has been considerable confusion in the past between this

species and Ampelisca excavata K. H. Barnard, 1925. K. H. Barnard’s descrip-

tion of A. excavata was based on a single specimen, and when larger specimens

of an apparently similar form were collected he ascribed these to the same species,

concluding that his original specimen must have been abnormal (K. H. Barnard

1955).

THE AMPHIPODA OF SOUTHERN AFRICA Hf

Subsequently it has been found that these two batches of material represent

distinct species. K. H. Barnard’s original A. excavata has been redescribed in

great detail by Gray & J. L. Barnard (1970) while his 1955 material is hereby

renamed Ampelisca acris n. sp.

As can be readily appreciated by comparing Figure 3 with the illustrations

in Gray & J. L. Barnard (1970), the two forms are quite distinct. A. excavata

is altogether a stouter species with a bifurcation at the tip of article 2 of pereiopod

5, a reduced inner ramus of uropod 1 and heavily chitinized rami on uropods

2 and 3. As well as differing in the structural features mentioned above, the two

species can be distinguished by their modes of life, A. acris being a free-living

tube builder, whereas A. excavata appears to be restricted to cirripede burrows

in the shells of large molluscs such as Turbo sarmaticus Linn. and Haliotis

midae Linn.

Material: SCD 160G(2); LIZ 17E(4); MB 50Q(3), MB 54X(1), MB 66V(1).

Ampelisca anisuropa (Stebbing, 1908)

Byblis anisuropus Stebbing, 1908a: 72, pl. 10. K. H. Barnard, 1955: 82, fig. 40B.

Ampelisca anisuropa: Griffiths, 1974a: 220.

Records: SCD 104X(1), SCD 173X(1), SCD 188D(2), SCD 204K(3), SCD

211Z(1), SCD 232C(4), SCD 321Q(2), SCD 392U(2); SST 5J(1), SST 9F(1),

SST 16E(2); 33°S/28°E/86 m (Stebbing 1908a).

Distribution: Endemic, Natal to west coast of South Africa.

Ampelisca anomala Sars, 1882

Ampelisca anomala: Sars, 1895: 178, pl. 62, fig. 2.

Records: SCD 24J(1), SCD 41L(4); LIZ 29Q; MB 57E(1), MB 50R(1).

Diagnosis: Head with post-antennal corner acute; two pairs of eyes with large

corneal lenses, the lower pair directed laterally; antennae fairly long, antenna

1 half as long as body, 2 slightly shorter than body; none of coxae with distal

teeth; article 2 of pereiopod 5 distally rounded, 3 longer than 4, 4 not produced

distally, 6 equal to 4 plus 5, 7 of moderate size; third pleonal epimeron quadrate

postero-distally; pleon segment 4 with a distinct triangular dorsal carina; telson

80% cleft, a single spinule at apex of each lobe.

Distribution: Scandinavia, South Africa.

Ampelisca brevicornis (Costa, 1853)

Ampelisca brevicornis: Reid, 1951: 204—210, figs 9-15. Kaim Malka, 1969: 928-932, pls 1-6.

Records: SCD 24K(9), SCD 95C(1), SCD 103W(1), SCD 122E(1), SCD 148K(4),

SCD 160F(1), SCD 198E(1), SCD 202E(1), SCD 219F(3), SCD 222H(15),

SCD 225H(12), SCD 244B(155), SCD 248H(20), SCD 257M(5), SCD 276M(3),

Dip ANNALS OF THE SOUTH AFRICAN MUSEUM

SCD 278J(3), SCD 315V(102), SCD 329V(2), SCD 338L(5), SCD 343T(1),

SCD 349R(1), SCD 353M(1), SCD 366T(1), SCD 368V(1), SCD 370K(6),

SCD 379N(1), SCD 381P(1), SCD 391F(4); SST 29E(42), SST 32F(5), SST

34F(14), SST 40S(10), SST 41N(1), SST 45D(8), SST 62M(2), SST 65B(4),

SST 68N(2); 34°S/22°E/2m (K. H. Barnard 1916).

Distribution: Cosmopolitan.

Ampelisca chiltoni Stebbing, 1888

Ampelisca chiltoni: J. L. Barnard, 1961: 61, fig. 31.

Records: SCD 160E(1), SCD 232A(2), SCD 262G(1), SCD 356N(22), SCD

359A(1), SCD 392X(15); SST 1D(10), SST 5B(24), SST 9C(16), SST 16F(8),

SST 47H(1); LIZ 32W(11); 32°S/28°E/86 m (Stebbing 1918).

Distribution: Australia, New Zealand, South Africa.

Ampelisca diadema (Costa, 1853)

Ampelisca diadema: Chevreux & Fage, 1925: 82, fig. 74.

Records: LIZ 32X(11).

Distribution: Europe, southern Africa.

Ampelisca fusca Stebbing, 1888

Ampelisca fusca Stebbing, 1888: 1052, pl. 105.

Records: SCD 181J(26), SCD 204L(20), SCD 235X(3), SCD 253G(1), SCD

269S(1), SCD 273E(1), SCD 300P(34), SCD 302T(1), SCD 321R(1); SST

19A(230), SST 24A(260), SST 34P(7).

Distribution: Endemic, Mocambique to South West Africa.

Ampelisca natalensis K. H. Barnard, 1916

Ampelisca natalensis K. H. Barnard, 1916: 137, pl. 26, fig. 7.

Records: LIZ 19P(1), LIZ 290(1).

Distribution: Endemic, Natal to Port Elizabeth.

Ampelisca palmata K. H. Barnard, 1916

Ampelisca palmata K. H. Barnard, 1916: 136, pl. 28, figs 30-31.

Records: SCD 41M(6), SCD 93M(1), SCD 115F(2), SCD 141V(2), SCD

188C(10), SCD 192V(1250), SCD 193W(2), SCD 198F(3), SCD 204M(14),

SCD 219G(7), SCD 232B(1), SCD 235W(17), SCD 248J(70), SCD 321P(1),

THE AMPHIPODA OF SOUTHERN AFRICA 273

SCD 343R(47), SCD 356P(12), SCD 376L(7), SCD 381Q(1), SCD 383T(2),

SCD 392Y(1); SST 56S(1); MB 10Q(1); ZZ 3U; 32°S/28°E/94 m (K. H. Barnard

1916).

Distribution: Senegal to Mocgambique.

Ampelisca spinimana Chevreux, 1887

Ampelisca spinimana: Chevreux & Fage, 1925; 81, fig. 73.

Records: SST 45J(4); KNY 245H(1); BMR 23T(1).

Distribution: Europe, West and South Africa.

Byblis gaimardi (Kroyer, 1846)

Byblis gaimardi: Mills, 1971: 367-370, figs 6A, 7.

Records: SCD 300R(4), SCD 302X(1); 32°S/28°E/94 m (K. H. Barnard 1916).

Distribution: Arctic, North Atlantic, ? Pacific, South Africa.

Family Amphilochidae

Cyproidea ornata Haswell, 1880

Cyproidea ornata: J. L. Barnard, 1972a: 21, figs 4, 5.

Records: SCD 216D(13); J 11F; K 8P; E 223; L 504; Port Alfred, East London

(K. H. Barnard 1940).

Distribution: Indo-Pacific, extending to South West Africa.

Gitanopsis mariae Griffiths, 1973

Gitanopsis mariae Griffiths, 1973: 275, fig. 4.

Records: SST 11X(5), SST 24Q(1).

Distribution: Endemic, Mogambique to Still Bay.

Gitanopsis pusilla K. H. Barnard, 1916

Gitanopsis pusilla K. H. Barnard, 1916: 144.

Records: SCD 55K(1), SCD 74J(2), SCD 173Y(28), SCD 198Y(3), SCD 392Z(1);

SST 32M(1), SST 37W(4); HAV 3C(1); SS 55C; Still Bay (K. H. Barnard 1940),

Distribution: Southern Atlantic, southern Indian Ocean.

Hoplopleon medusarum K. H. Barnard, 1932

Hoplopleon medusarum K. H. Barnard, 1932: 105, fig. 54.

Records: SCD 338K(1); SST 16B(3).

Distribution: Endemic, south coast of South Africa to South West Africa.

274 ANNALS OF THE SOUTH AFRICAN MUSEUM

Family Ampithoidae

Ampithoe africana K. H. Barnard, 1925

Ampithoe africana K. H. Barnard, 1925: 361.

Records: KNY 166F; Y 12G; East London (K. H. Barnard 1925).

Distribution: Endemic, Natal to Knysna.

Ampithoe falsa K. H. Barnard, 1932

Ampithoe brevipes: K. H. Barnard, 1916: 255, pl. 28, fig. 34.

Ampithoe falsa: Ruffo, 1969: 57, figs 18-20.

Records: LLL 6L; SS 55H; Still Bay (K. H. Barnard 1940).

Distribution: Indian Ocean.

Ampithoe ramondi (Audouin, 1826)

Ampithoe intermedia: Stebbing, 1910a: 462.

Ampithoe vaillanti: K. H. Barnard, 1916: 253.

Ampithoe ramondi: J. L. Barnard, 1970b: 50, figs 18, 19.

Records: SCD 20U(1); MB 10S(1), MB 40L(3), MB 58Q(1), MB 59F(1);

LIZ 40F; L 493B; 33°S/26°E/18-29 m (Stebbing 1910a).

Distribution: Circumtropical.

Cymadusa filosa Savigny, 1818

Grubia australis K. H. Barnard, 1916: 258.

Cymadusa australis: K. H. Barnard, 1940: 480.

Cymadusa filosa: J. L. Barnard, 1955: 29, fig. 15.

Records: KNY 6L, KNY 11H, KNY 30A, KNY 122A, KNY 128A, KNY

139C(9), KNY 162A, KNY 179D(10), KNY 245G(4), KNY 283K(3); Knysna

Lagoon (K. H. Barnard 1940).

Distribution: Circumtropical.

Macropisthopus stebbingi K. H. Barnard, 1916

Macropisthopus stebbingi K. H. Barnard, 1916: 260, pl. 28, figs 15-17.

Records: ZZ3R; SS 55K; Port Elizabeth (K. H. Barnard 1916); Still Bay

(K. H. Barnard 1940).

Diagnosis: This species is the only representative of the genus, which is diagnosed

as follows: antenna 1 without accessory flagellum; mandible with palp; both

THE AMPHIPODA OF SOUTHERN AFRICA 275

gnathopods very feebly chelate; article 6 of pereiopods 3-5 scarcely widened;

pereiopod 5 greatly enlarged, flattened; outer ramus of uropod 3 bearing two

large hooks.

Distribution: South coast of South Africa.

Family Cheluridae

Chelura terebrans Philippi, 1839

Chelu;z terebrans: Chevreux & Fage, 1925: 371, figs 379, 380.

Records: Port Elizabeth harbour (Stebbing 1910a).

Diagnosis: Head with large frontal process; pleon segment 1 with large medio-

dorsal posteriorly directed process; urosome segments fused; uropods differing

radically from one another in size and form; uropod 2 with short subequal rami,

peduncle with large marginally setose winglike lobe; uropod 3 enormous, outer

ramus very large, inner ramus small.

Distribution: Cosmopolitan.

Family Colomastigidae

Colomastix keiskama n. sp.

Fig. 4

Description of male (4 mm): Head with anterior keel projecting between the

antennae, eyes of moderate size, brown (in alcohol), lateral cephalic lobes

evenly rounded; antenna | lacking spines, articles 1-3 each with a pair of ventral

keels distally produced into small teeth, flagellum of a single terminally setose

article; antenna 2 smooth, slightly shorter than 1; mandible (Fig. 4B) cut into

five strong teeth, the first terminally bifurcate; maxilliped (Fig. 4C) with inner

plates coalesced, outer plates each bearing a single terminal seta.

Coxae 1-6 smoothly rounded, oval, coxa 6 produced posteriorly into a

semi-acute point (Fig. 4A); articles 5 and 6 of gnathopod 1 subequal, article 6

terminating acutely in about five setae; article 2 of gnathopod 2 faintly crenulate

anteriorly, article 5 unusually large, bearing a wide setose posterior lobe, article 6

not more than 1,5 times the size of 5, as wide as long, palm about equal to hind-

margin, defined by three small cusps and bearing two subequal teeth near

finger hinge, dactyl equal to palm; pereiopod 3 smaller than 4 or 5, article 2

oval.

Pleonal epimeron 1-3 broadly rounded; uropod 1 with inner ramus ter-

minating in a strong blade-like spine about 50 per cent length of body of ramus;

uropods 2 and 3 projecting about equally and well beyond the tip of uropod 1,

inner ramus of each marginally the longer and with its upper margin finely

serrate; telson smoothly rounded (Fig. 4D).

276 ANNALS OF THE SOUTH AFRICAN MUSEUM

“ApS

Fig. 4. Colomastix keiskama n. sp.

Male, 4 mm: A—lateral aspect; B—mandible; C—maxilliped; D—fused urosomites 2 and 3 and telson in dorsal view.

Holotype: SAM A13208, male, 4 mm.

Type locality: SCD 179J, 33°59’S/25°41’E, 24 November 1960, depth 4-11 m,

substrate rocky.

Relationships: The second gnathopod and first uropod are of most unusual

structure and immediately distinguish this species from its relatives. Other

species with equal rami of uropod 3 and an unsculptured telson include C. pusilla

THE AMPHIPODA OF SOUTHERN AFRICA O77

Grube and C. simplicauda Nicholls but their second gnathopods are quite

different from those of C. keiskama n. sp.

Material: SCD 179J, a single male.

Colomastix pusilla Grube, 1864

Colomastix pusilla: J. L. Barnard, 1971: 55, fig. 24.

Records: SCD 159U(20).

Distribution: Cosmopolitan in tropical and temperate seas.

Family Corophiidae

(Revised J. L. Barnard 19725)

Aora typica Kroyer, 1845

Aora typica: J. L. Barnard, 19695; 148, fig. 63.

Records: SCD 62C(1), SCD 118N(1), SCD 135D(1), SCD 148J(2), SCD 159Y(3),

SCD 181V(4), SCD 198H(2), SCD 208G(1), SCD 225J(4), SCD 232D(\),

SCD 244K(2), SCD 257H(5), SCD 276Q(2), SCD 287C(2), SCD 310H(1),

Sem ZIG): SCD 326K(1), SCD 338HG), SCD 339R(Q), SCD 353PQ@),

SCD 368V(1), SCD -379M(1), SCD 381S(6), SCD 392ZQ); SST 16P(1),

som S2B(1), SST 3431), SST 37UG); LIZ 29RQ); TF 133; SS 55G.

Distribution: Cosmopolitan.

Camacho bathyplous Stebbing, 1888

Camacho bathyplous: J. L. Barnard, 1961: 115, figs 81, 82.

Records: 33°S/28°E/86 m (Stebbing 1908a).

Diagnosis: This genus is monotypic, the diagnosis being: antenna | elongate,

greatly exceeding antenna 2, article 1 much longer than 3, accessory flagellum

multi-articulate; mandibular palp 3-articulate; coxae short, serially discon-

tinuous; gnathopods subchelate, uropods biramous, the third with inner ramus

30% length of outer.

Distribution: Indo-Pacific.

Cerapus tubularis Say, 1818

Cerapus abditus: K. H. Barnard, 1916; 271.

Cerapus tubularis: J. L. Barnard, 1962: 61, figs 27, 28.

Records: SCD 41K(3), SCD 60H(4), SCD 81B(1), SCD 159Z(1), SCD 244A(25),

Sep 2575(3), SCD 262M(1), SCD 267LQ), SCD- 312P(5), SCD 338E(1),

SCD 342F(4), SCD 343V(4), SCD 353C(20); SST 32E(1); LIZ 13U().

Distribution: Cosmopolitan in warm and temperate seas.

278 ANNALS OF THE SOUTH AFRICAN MUSEUM

Cheiriphotis megacheles (Giles, 1885)

Cheiriphotis megacheles: J. L. Barnard, 1962: 17, fig. 4.

Records: SCD 20M(10), SCD 20L(2), SCD 34D(1), SCD 34G(1), SCD 60J(20),

SCD 62A(C), SCD 64B(C), SCD 81D(2), SCD 95J(18), SCD 124U(4), SCD

124V(1), SCD 189R(38), SCD 192Y(1), SCD 257G(101), SCD 278K(2),

SCD 282S(10), SCD 283F(18), SCD 285S(4), SCD 312Q(7), SCD 329W(2),

SCD 332S(15), SCD 338G(181), SCD 339Q(9), SCD 353F(14), SCD 379J(10),

SCD 383V(1), SCD 391K(1); SST 45F(2); LIZ 3Z(12), LIZ 31S(1); MB 4U(4),

MB 5H(4), MB 28G(8), MB 33N(1), MB 45E(1), MB 46F(1), MB 59H(1)

MB 71K(2), MB 73M(7), MB 75K(4), MB 77Q(3), MB 80D(6).

Distribution: Indo-Pacific.

Chevalia aviculae Walker, 1904

Chevalia aviculae: J. L. Barnard, 1971: 88, fig. 42.

Records: SCD 62D(1), SCD 99N(4), SCD 181U(2), SCD 2445(5), SCD 302V(1),

SCD 304R(18), SCD 308J(1), SCD 311J(2), SCD 366X(1); SST 16A(9),

SST 19H(1), SST 24L(6), SST 47E(1); 34°S/23°E/230 m (K. H. Barnard 1916).

Distribution: Cosmopolitan in tropical and temperate seas.

Concholestes armatus n. sp.

Figs 5, 6

Description of male (6 mm): Ocular lobes of head strongly produced obliquely

downwards, eyes small, black; antennae very large (Fig. 5A), 1 slightly shorter

than body, articles 1 and 2 subequal, longer than 3, flagellum 12-articulate;

antenna-2 longer than body, very sturdy, used to drag the animal and its

Dentalium shell abode across the sea floor, flagellum of one long and one short

article; mandible (Fig. 6A) bearing large molar composed of rows of sharp

tubercles, palp uni-articulate, bearing long plumose setae; maxilla 1 (Fig. 6B)

with bi-articulate palp, article 2 terminating in six strong serrate spines, outer

plate also ending in six strong serrate spines, inner plate smooth; plates of

maxilla 2 subequal, the inner with an oblique row of 16 setae medially (Fig. 6C),

both plates terminally setose; maxilliped (Fig. 6D) with 4-articulate palp, inner

plate with three terminal blade spines, outer plate with proximal spines blade-

like, the more distal ones becoming longer and more slender.

Coxae | and 2 produced forwards, marginally setose; gnathopod | simple

(Fig. 5B), article 5 longer than 6, dactyl extremely large, bearing seven strong

spines along its posterior margin; gnathopod 2 (Fig. 5C) subchelate, article 2

broad, six very strong spines and one small one forming a comblike row along

anterior margin, article 3 bearing two anterior spines, article 6 longer than 5,

THE AMPHIPODA OF SOUTHERN AFRICA 279

> ae

SSS Saas

od CRE,

Fig. 5. Concholestes armatus n. sp

Male, 6mm: A—lateral aspect; B—gnathopod 1; C—gnathopod 2; D—pereiopod 1 ; E—pereiopod 3; F—pereiopod 5.

widest at its base, palm spinose, dactyl shorter than palm, bearing a distal

spine; coxa 3 anteriorly lined by a row of short spines and setae, posterior

margin lined by long plumose setae; article 2 of pereiopod 1 subcircular (Fig.

5D), article 5 posteriorly covered in short thick spines; coxa 4 triangular, not

spinose; pereiopod 2 like 1; pereiopod 3 small, directed posteriorly (Fig. 5E),

article 5 bilobed, outer lobe covered in minute spines, inner lobe bearing many

large hooked spines; pereiopod 4 similar to 3 but somewhat larger; pereiopod

5 (Fig. 5F) elongate, article 2 strongly setose.

280 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 6. Concholestes armatus n. sp.

Male, 6 mm: A—mandible; B—maxilla 1; C—maxilla 2; D—maxilliped; E—dorsal view of

urosome; F—ventral view of uropod 1; G—uropod 2 as seen laterally.

Pleonal epimera rounded, setose; urosome (Fig. 6E) reduced, tucked

between pereiopods, only two segments present; segment | bearing a pair of

large biramous uropods, both peduncle and rami heavily spinose (Fig. 6E, F),

inner ramus 50% length of outer; second urosome segment less than half length

of first, bearing a pair of laterally flattened uropods represented by a distally

bilobed peduncle (Fig. 6E, G); telson very large, squamous, laterally strongly

spinose.

Female: Ovigerous at 4 mm: the eggs are restrained by the pereiopods which

angle in ventrally to forrn a tunnel beneath the body, the brood lamellae are

THE AMPHIPODA OF SOUTHERN AFRICA 281

reduced to narrow lobes. In one case newborn young were found to occupy the

narrow end of a Dentalium shell in which a large female was living.

Holotype: SAM A13207, male, 6 mm (13 mm including antennae).

Type locality: SST 1B, 35°22’S/22°31’E, 20 June 1972, depth 200 m, substrate

coarse khaki sand.

Relationships: This species closely resembles C. dentali Giles in its habits and

in its unusually strong antennae and reduced third uropods. Unfortunately,

however, Giles’s description of the urosome of his species is unclear and he

fails to describe its mandibular palp. The urosome is described as follows:

“Ath abdominal appendage biramous; 6th blunt, rounded, without rami, nearly

hidden beneath the squamous telson . . . of the fifth abdominal appendage I

have been unable to obtain a satisfactory view, it is small and its peduncle is

very short, though of considerable width. The ramus appears to be single and

rounded.’

Despite his mention of a third uropod Giles’s figure of the urosome (also

reproduced by J. L. Barnard, 19695) shows only two urosome segments and two

pairs of uropods. Assuming Giles’s verbal description of this unexpected and

difficult to observe urosome to be erroneous, I have not erected a new genus

for my species, especially since the chances of two distantly related forms

occupying an identical niche seem remote. Should re-examination of C. dentali

reveal a third urosomite or a mandibular palp of more than 1 article, this

would necessitate the erection of a new genus for C. armatus.

Material: SST 1B(4), SST 5A(1), SST 11N(7).

Corophium acherusicum Costa, 1857

Corophium acherusicum: J. L. Barnard, 1971: 59, figs 17, 26.

Records: LIZ 1K(2); MB 87N(1).

Distribution: Cosmopolitan in tropical and temperate seas.

Corophium triaenonyx Stebbing, 1904

Corophium triaenonyx Stebbing, 1904: 25, pl. 6A.

Records: SUN 5E(1); BMR 25G(2); HAV 3A(A), HAV 5B(2), HAV 7N(C),

HAV 10D(C), HAV 18K(C); KNY 112B, KNY 139D, KNY 160A, KNY 166E,

KNY 176D, KNY 181B, KNY 184B, KNY 191E; GBR 23D, GBR 37B(14),

GBR 46A(2); STJ 15J(P), STJ 16G(C), STJ 26B(C), STJ 27L(2), STJ 31T(A),

STJ 32B(C); Keurbooms River, Plettenberg Bay, Knysna (K. H. Barnard 1940).

Distribution: Atlantic, Mediterranean, Indian Ocean.

Ericthonius brasiliensis (Dana, 1853)

Ericthonius brasiliensis: J. L. Barnard, 1971: 61, fig. 17E.

Records: LIZ 40E(10); MB 40V(1), MB S5OW(C), MB 87B(4); KNY 139E;

L 458B; 32°S/26°E/18-29 m (Stebbing 1910a).

282 ANNALS OF THE SOUTH AFRICAN MUSEUM

Distribution: Cosmopolitan in tropical and temperate seas.

Gammaropsis afra Stebbing, 1888

Eurystheus afer: Stebbing, 1910a: 461.

Gammaropsis afra: J. L. Barnard, 1970b: 170, fig. 108.

Records: 33°S/28°E/86 m, 33°S/26°E/18-29 m (Stebbing 1908a).

Distribution: Circumtropical.

Gammaropsis atlantica Stebbing, 1888

Eurystheus atlanticus: Stebbing, 1910a: 461.

Gammaropsis atlantica: J. L. Barnard, 1971: 91, figs 43-45.

Records: SCD 10S(4), SCD 34C(3), SCD 55D(1), SCD 81A(1), SCD 93P(1),

SCD 95A(50), SCD 102D(3), SCD 103T(1), SCD 128Q(3), SCD 172V(1),

SCD 181X(1), SCD 185J(1), SCD 204R(1), SCD 208K(5), SCD 222L(1),

SCD 269R(2), SCD 280K(1), SCD 300Q(5), SCD 302P(26), SCD 304Q(1),

SCD 319X(4), SCD 321N(3), SCD 324G(3), SCD 328G(1), SCD 345S(2),

SCD 352A(2), SCD 353A(24), SCD 392R(13); SST 5G(4), SST 16D(4),

SST 24C(1); LIZ 37L(2), LIZ 40B(2); MB 50X(1), MB 61U(C), MB 66P(1),

MB 87D(5), MB 86C(1); 33°S/28°E/86 m, 33°S/26°E/18-29 m (Stebbing 1908a).

Distribution: Circumtropical.

Gammaropsis holmesi (Stebbing, 1908)

Eurystheus holmesi Stebbing, 1908a: 85, pl. 14A.

Eurystheus semidentatus K. H. Barnard, 1916: 250, pl. 28, figs 13, 14.

Records: SCD 62E(2), SCD 135K(1); LIZ 40H(2); 33°S/26°E/18—29 m (Stebbing

1908a).

Distribution: Endemic, Natal to Saldanha Bay.

Grandidierella bonnieroides Stephensen, 1948

Grandidierella bonnieri: Ledoyer, 1967: 137, fig. 283A. Griffiths, 1973: 283; 19746: 228.

(non) Grandidierella bonnieri Stebbing, 19085: 120, pl. 16.

Grandidierella bonnieroides: Myers, 1970: 141, figs 1, 2.

Records: STJ 6A(A).

Distribution: Indian Ocean, Caribbean.

Remarks: Myers (1970) re-examined Stebbing’s type material of G. bonnieri

and found it to have only a single distal tooth on article 5 of gnathopod 1, as

compared with the three distal teeth present in the material of Ledoyer and

THE AMPHIPODA OF SOUTHERN AFRICA 283

Griffiths. These records are thus transferred to G. bonnieroides Stephensen.

Myers further points out that different populations of G. bonnieroides from

differing locations show variation in sternal armature, and that this factor is

thus not of taxonomic significance.

Grandidierella chelata K. H. Barnard, 1951

Grandidierella chelata K. H. Barnard, 1951: 708, fig. 7.

Records: HAV 7H(A), HAV 17A(A); STJ 14A(A), STJ 15G(A), STJ 16H(P),

STJ 27J(P).

Distribution: Endemic, Port St. Johns to South West Africa.

Grandidierella lignorum K. H. Barnard, 1935

Grandidierella lignorum K. H. Barnard, 1935: 300, fig. 14.

Records: BMR 23S(4), BMR 25J(1), BMR 26U(1); HAV 3B(C), HAV 9A(5),

HAV 13K(A); KNY 101A, KNY 184A; GBR 46C(2); BRE 51M(3), BRE

52M(7), BRE 71M(C), BRE 77H(3), BRE 81A(C); STJ 8A, STJ 14B(P), STJ

ISH(P), STJ 24K(P), STJ 27K(P); Keurbooms River, Plettenberg Bay (K. H.

Barnard 1940).

Distribution: Estuaries around the Indian Ocean.

Lembos hypacanthus K. H. Barnard, 1916

Lembos hypacanthus K. H. Barnard, 1916: 237, pl. 28, figs 5, 6.

Lembos hirsutipes (non Stebbing 1895): K. H. Barnard, 1951: 706.

Records: KNY 13G, KNY 139B, KNY 166G, KNY 171F.

Distribution: Endemic, Natal to South West Africa.

Neomicrodeutopus nyala n. sp.

Fig. 7

Description of male (4 mm): Head anteriorly truncate, eyes circular with the

centre black; mandible with 3-articulate palp, articles 1 and 3 subequal, shorter

than 2, molar large, lacinia mobilis present (Fig. 7E), spine row of three

spines; inner plates of maxilliped (Fig. 7F) with three apical blade-spines, outer

plate with seven blade-spines along inner margin, the spines becoming pro-

gressively longer distally, palp 4-articulate, the terminal article bearing three

strong setae; (antennae missing).

Coxae short, not touching serially; gnathopod | (Fig. 7B) powerful, article 5

postero-distally produced into two teeth separated by a semicircular concavity,

article 6 much narrower than 5, bearing a small bump on posterior margin,

284 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 7. Neomicrodeutopus nyala n. sp.

Male, 4 mm: A—lateral aspect; B—gnathopod 1; C—gnathopod 2 left side;

D-—article 2 of gnathopod 2 right side; E—mandible; KF—maxilliped;

G—uropod 3.

THE AMPHIPODA OF SOUTHERN AFRICA 285

7 shorter than 6; article 2 of gnathopod 2 (Fig. 7C) anteriorly produced into a

number of projections (in this individual four fairly regular teeth on one side

and three teeth on the other), the distal tooth the largest, article 3 anteriorly

produced into an elongate lobe, article 5 longer than 6, palm oblique, not

defined, dactyl terminating in a large spine and three setae; pereiopod | stout,

article 2 regularly crenulate anteriorly; (pereiopods 2—5 missing).

Pleonal epimera smoothly rounded, somewhat lobed postero-distally;

uropods | and 2 biramous, rami equal, terminally strongly spinose, uropod |

with a terminal peduncular spine projecting between the rami; uropod 3

(Fig. 7G) uniramous, ramus twice length of peduncle, uni-articulate, three long

setae at its apex; telson quadrangular, a dorsal tubercle and two or three short

setae on each corner.

Holotype: SAM A13069, male, 4 mm.

Type locality: SST 34Q, 34°40’S/21°39’E, 21 June 1972, depth 80 m, substrate

coarse shelly sand.

Relationships: The unusual armature of article 2 of gnathopod 2 immediately

distinguishes this species from the other three members of the genus. In addition

N. nyala n. sp. differs from N. makaka J. L. Barnard in the narrower article 6

of gnathopod 1 and the uni-articulate ramus of uropod 3, and from N. cabindae

Schellenberg in the shape of article 5 of gnathopod 1. N. elongata (Chevreux)

has more produced optic lobes and a much longer peduncle of uropod 3 than

N. nyala n. sp.

Material: SST 34Q, 2 gd.

Photis dolichommata Stebbing, 1910

Photis dolichommata Stebbing, 1910b: 609, pl. 55B.

Records: SCD 34E(6), SCD 41P(4), SCD 81C(2), SCD 100P(2), SCD 102H(5),

SCD 103U(1), SCD 181S(8), SCD 204Q(5), SCD 208H(2), SCD 235Y(1),

SCD 249U(1), SCD 295A(1), SCD 353B(260); SST 1A(3), SST 16T(1);

34°S/22°E/230 m (K. H. Barnard 1916).

Diagnosis: Ocular lobes very long, eyes oval, large; palm of gnathopod | male

oblique, not excavate, dactyl finely serrate; article 2 of gnathopod 2 male not

lobed antero-distally, palm slightly oblique, defined by a fairly strong tooth and

with two small processes along its length; inner ramus of uropod 3 30% length

of outer, outer ramus with a minute second article.

Distribution: Australia, South Africa.

Photis kapapa J. L. Barnard, 1970

Photis kapapa J. L. Barnard, 1970b: 192, figs 124, 125.

Records: SCD 282V(3), SCD 287E(4); SST 16U(1).

Distribution: Hawaii, east coast of southern Africa.

286 ANNALS OF THE SOUTH AFRICAN MUSEUM

Photis longimanus Walker, 1904

Photis longimanus: K. H. Barnard, 1916: 244. Sivaprakasam, 1969: 567, fig. 8.

Records: SCD 192Z(40), SCD 204Z(1), SCD 244M(10), SCD 248K(3), SCD

257P(4), SCD 338Q(6), SCD 339T(4), SCD 368S(5).

Distribution: Indian Ocean, extending to South West Africa.

Photis uncinata K. H. Barnard, 1932

Photis longicaudata: K. H. Barnard, 1916: 243, pl. 28, fig. 26.

Photis uncinata K. H. Barnard, 1932: 223, fig. 138.

Records: SCD 24N(4), SCD 60L(2), SCD 83D(3), SCD 95D(30), SCD 99L(1),

SCD 103V(1), SCD 122F(10), SCD 131Y(4), SCD 141T(1), SCD 146A(24),

SCD 148M(1), SCD 151G(7), SCD 159X(3), SCD 181W(2), SCD 184E(2),

SCD 188A(27), SCD 189A(135), SCD 216G(2), SCD 244P(5), SCD 302U(1),

SCD 319Z(1), SCD 321U(2), SCD 343Z(2), SCD 345Z(6), SCD 353J(38),

SCD 370S(3), SCD 379Q(1), SCD 392Z(8); SST 19B(6), SST 24D(3), SST 29F(4),

SST 32G(1), SST 34K(3), SST 40U(2); 33°S/28°E/83 m, 34°S/23°E/80 m (K. H.

Barnard 1916).

Distribution: Endemic to South Africa.

Siphonoecetes dellavallei Stebbing, 1893

Siphonoecetes dellavallei: Chevreux & Fage, 1925: 361, fig. 369.

Records: SCD 135J(3), SCD 198M(5), SCD 236C(3), SCD 285R(2), SCD

300W(1), SCD 338C(450), SCD 339V(7), SCD 368Q(30), SCD 383U(10),

SCD 3848(9); SST 24M(1), SST 32D(1), SST 34N(1), SST 45A(33), SST 65G(1),

SST 67Q(2), SST 73F(2); MB 5G(13), MB 34L(5).

Distribution: Mediterranean, southern Africa.

Siphonoecetes orientalis Walker, 1904

Siphonoecetes orientalis Walker, 1904: 294, pl. 7, fig. 49. K. H. Barnard, 1916: 270.

Records: SCD 95F(6), SCD 96W(1), SCD 120B(1), SCD 192Z(6), SCD 237H(1);

32°S/28°E/95 m, 34°S/26°E/116 m (K. H. Barnard 1916).

Distribution: Tropical Indo-Pacific.

Family Dexaminidae

Atylus granulosus (Walker, 1904)

Atylus granulosus: Ledoyer, 1967: 127, fig. 8.

Records: SCD 24Y(1); SST 37S(2), SST 47G(1), SST 60M(1); LIZ 31P(1).

Distribution: Indian Ocean.

THE AMPHIPODA OF SOUTHERN AFRICA 287

Atylus guttatus (Costa, 1851)

Nototropis guttatus: Chevreux & Fage, 1925; 194, figs 201-203.

Records: SCD 285L(5), SCD 332V(34), SCD 339W(1), SCD 383N(141).

Distribution: Europe, West and South Africa.

Atylus homochir Haswell, 1885

Atylus homochir: Stebbing, 1888: 908-913, pl. 74.

Records: 33°S/26°E/18-29 m (Stebbing 1910a).

Diagnosis: Body dorsally carinate but carinae produced into teeth only on

pereon segment 7 and pleon segments 1-3; pleon segment 4 with two dorsal

teeth, the hind one the larger; composite pleon segment 5 and 6 with a posterior

tooth; article 2 of pereiopod 3 not at all produced postero-distally; pleonal

epimera 1-3 each produced into a minute point postero-distally.

Distribution: Australia, South Africa.

Atylus swammerdami (Milne-Edwards, 1830)

Paratylus swammerdami: Sars, 1895: 463, pl. 163.

Atylus swammerdami: Chevreux & Fage, 1925: 195, fig. 204.

Records: MB 20V(1), MB 28F(4), MB 69S(1), MB 82B(17).

Distribution: Europe, West and South Africa.

Polycheria atolli Walker, 1905

Polycheria atolli: Ledoyer, 1972: 205, pl. 27.

Records: SCD 10R(5), SCD 55F(6), SCD 160A(11), SCD 172U(54), SCD

173T(8), SCD 181P(30), SCD 244H(51), SCD 262N(1), SCD 366P(1), SCD

379L(1), SCD 388D(1); SST 16L(1); LIZ 17F(3), LIZ 37N(1); MB 54V(C),

MB 69Q(3), MB 77R(1), MB 86B(1), MB 87M(2); L 410; VV 2H; Still Bay

(K. H. Barnard 1940).

Distribution: Southern oceans, extending to tropical Indian Ocean.

Family Eusiridae

Eusirus minutus Sars, 1893

Eusirus minutus: Sars, 1895: 419, pl. 149, fig. 2.

Records: SST 16Q(1).

Diagnosis: Pereon segment 7 and pleon segments 1 and 2 dorsally toothed;

article 6 of gnathopods 1 and 2 attached to upper distal corner of article 5

(‘eusirid’); third pleonal epimeron postero-distally rounded, lower posterior

288 ANNALS OF THE SOUTH AFRICAN MUSEUM

margin finely serrate; article 6 of pereiopods 3-5 less than twice as long as

article 2; telson evenly tapering, less than 20% cleft, apices divergent.

Distribution: Norway, South Africa.

Eusiroides monoculoides (Haswell, 1880)

Eusiroides monoculoides: J. L. Barnard, 1964: 221, fig. 1.

Records: SCD 59C(2), SCD 181Z(3), SCD 253E(2), SCD 300N(18), SCD

312R(9), SCD 353Q(4), SCD 366Y(2); LIZ 29P(1); 32°S/28°E/170 m, 33°S/28°E/

120 m, 34°S/25°E/137 m (K. H. Barnard 1916).

Distribution: Circumtropical.

Paramoera bidentata K. H. Barnard, 1932

Paramoera bidentata K. H. Barnard, 1932: 211, figs 118m, 129.

Records: Still Bay (K. H. Barnard 1940).

Distribution: Endemic, Still Bay to South West Africa.

Paramoera capensis (Dana, 1853)

Paramoera capensis: K. H. Barnard, 1916: 183-186.

Paramoera schizurus Stebbing, 1918: 66, pl. 10.

Records: SCD 24Q(14), SCD 110U(1), SCD 285N(5), SCD 332U(22), SCD

338Q(3), SCD 339U(2), SCD 381V(1), SCD 391P(5); SST 16G(10), SST 47F(2),

SST 65D(2), SST 73J(1); LIZ 13Q(1), LIZ 29M(7), LIZ 32Y(1); MB 21D(1),

MB 28D(C), MB 32J(2), MB 33M(2), MB 38H(C), MB 57B(17), MB 59G(2),

MB 66Q(1), MB 70R(7), MB 71J(2), MB 73K(4), MB 82A(2), MB 87A(5);

KNY 13F, KNY 57B, KNY 139A, KNY 166C, KNY 171B(C), KNY 179B(1);

GBR 24H(7); BRE 135B(5); J 11J, Q 5J; LIZ 11Z; X 11B; K80; Y 12F; E 235;

ZZ 3M; T 3F; RR 4H; KN 2G; KKN 43G; SS 4L, SS 55D; East London,

Port Elizabeth (K. H. Barnard 1916); Still Bay (K. H. Barnard 1940).

Distribution: Atlantic and Indo-Pacific.

Family Gammaridae

Ceradocus rubromaculatus (Stimpson, 1855)

Ceradocus rubromaculatus: J. L. Barnard, 1972a: 220, fig. 129.

Records: SCD 55G(2), SCD 93L(3), SCD 110V(1), SCD 159S(2), SCD 181M(3),

SCD 262H(6), SCD 300M(8), SCD 353G(5), SCD 366N(4), SCD 388C(1);

SST 47B(8), SST 52P(1); LIZ 37K(1); MB 10R(2), MB 13N(1), MB 23G(2),

MB 40K(2), MB SOT(2), MB 54T(10), MB 66M(12), MB 77S(1), MB 84U(1);

THE AMPHIPODA OF SOUTHERN AFRICA 289

KNY 171A(1); QQ 4X; L 316; X 22E; E 229; ZZ 3T; T 3U; KN 43F; S 540;

33°S/25°E/45 m (Stebbing 1908a).

Distribution: Indo-Pacific.

Elasmopoides chevreuxi Stebbing, 1908

Elasmopoides chevreuxi Stebbing, 1908a: 82, pl. 39.

Records: 33°S/28°E/86 m (Stebbing 1908a).

Diagnosis: Accessory flagellum exceeding 20 articles; eyes extending almost

whole height of head; article 3 of mandibular palp as long as 2, article | simple;

inner plates of maxillae strongly setose medially; gnathopods subchelate; article

2 of pereiopods 3-5 strongly dentate posteriorly; uropod 3 not exceeding

uropod 1, rami broad, equal, the outer uni-articulate; lobes of telson basally

separate.

Distribution: The above record is the only one of this species to date.

Elasmopus japonicus Stephensen, 1932

Elasmopus spinimanus (non Walker, 1905): K. H. Barnard, 1925: 358.

Elasmopus japonicus: Sivaprakasam, 1968: 278, figs 3-5.

Records: Y 12K; E 230; Port Elizabeth (K. H. Barnard 1940).

Distribution: Indo-Pacific, extending to South West Africa.

Elasmopus pectenicrus Bate, 1862

Elasmopus pectenicrus: J. L. Barnard, 1970a: 125, figs 73, 74.

Records: J 11H; QQ 4Z; L 47S; X 22A; SS 55E; Still Bay, East London (K. H.

Barnard 1940).

Distribution: Cosmopolitan in tropical and temperate seas.

Eriopisa chilkensis (Chilton, 1921)

Niphargus chilkensis Chilton, 1921: 531, fig. 4.

Records: KNY 42A, KNY 81A.

Distribution: India, east coast of South Africa.

Eriopisa epistomata Griffiths, 1974

Eriopisa epistomata Griffiths, 1974a: 186, fig. 4.

Records: SCD 276 P(2); SST 40V(2), SST 41M(7).

Distribution: Endemic, south coast of South Africa to South West Africa.

290 ANNALS OF THE SOUTH AFRICAN MUSEUM

Eriopisella capensis (K. H. Barnard, 1916)

Eriopisa capensis K. H. Barnard, 1916: 187, pl. 27, figs 16-19.

Records: SCD 222M(6), SCD 228B(19), SCD 232J(10), SCD 356S(2).

Diagnosis: Antero-lateral angles of head rounded, eyes absent; mandibular

palp slender, article 3 shorter than 2; article 5 of gnathopod 1 not distally

widened, 6 oval, palm oblique, three times as long as hind margin; gnathopod

2 with article 5 triangular, wider than 6; pleonal epimera postero-distally

rounded; inner ramus of uropod 3 25% length of elongate outer ramus; telson

cleft to base, lobes dehiscent, each bearing 3—4 unequal spines.

Distribution: Endemic to south and west coasts of South Africa.

Maera boecki (Haswell, 1879)

Elasmopus boecki: K. H. Barnard, 1916: 199, pl. 27, figs 13, 14.

Maera boecki: K. H. Barnard, 1940: 460.

Records: S 54J; Port Elizabeth (K. H. Barnard 1916); Still Bay (K. H. Barnard

1940).

Diagnosis: Coxa | moderately produced forwards; palm of gnathopod 2 slightly

oblique, bearing four strong teeth, that closest to the finger hinge having one or

two accessory cusps, dactyl slender, closing within the defining tooth; posterior

margin of third pleonal epimeron entire; uropod 3 not exceeding uropod 1;

telson 60% cleft, lobes dehiscent, two setae arising from excavate tip of each

lobe.

Distribution: Australia, ‘South Africa.

Maera hamigera Haswell, 1880

Maera hamigera: K. H. Barnard, 1916: 196, pl. 27, fig. 11. J. L. Barnard, 1965: 507, fig. 16.

Records: SCD 262J(1), SCD 310G(20), SCD 343Z(1), SCD 373P(3), SCD

374K(5); 33°S/27°E/120 m (K. H. Barnard 1916).

Distribution: Indo-Pacific.

Maera inaequipes (Costa, 1851)

Meera inaequipes: J. L. Barnard, 1959a: 25, pl. S.

Records: SCD 10E(12), SCD 10T(3), SCD 24L(1), SCD 41N(10), SCD 55E(10),

SCD 102E(7), SCD 118P(1), SCD 181N(35), SCD 253D(1), SCD 300S(10),

SCD 302Q(1), SCD 353N(3), SCD 370R(1); SST 16J(2), SST 24E(1); MB

54Z(6), MB 86A(2); 32°S/28°E/141 m, 33°S/28°E/91 m (K. H. Barnard 1916).

Distribution: Cosmopolitan in tropical and temperate seas.

THE AMPHIPODA OF SOUTHERN AFRICA 291

Maera mastersi (Haswell, 1880)

Maera mastersi: Sivaprakasam, 1968: 36, fig. 1A-G. J. L. Barnard, 1972a: 226, fig. 132.

Records: SCD 34F(1), SCD 95G(1), SCD 159R(140), SCD 198Z(2), SCD

227P(1), SCD 269T(2), SCD 273H(2), SCD 321W(6), SCD 326L(2), SCD

366U(7); SST 16N(8), SST 19G(33).

Diagnosis: Coxa moderately produced forwards; gnathopod 2 much larger

than 1, palm defined by a small projection, proximal portion of palm smoothly

concave, covered by a dense fur of very short setae, distal third of palm pro-

duced into a small hump bearing two or three short spines; third pleonal epi-

meron bearing three posterior serrations; uropod 3 greatly exceeding 1 and 2;

telson cleft nearly to base, a small notch on inner margin of each apex bearing

a single seta.

Distribution: Indo-Pacific.

Remarks: The above specimens bear a close resemblance to those figured by

Sivaprakasam (1968), although his figures fail to show any ‘fur’ of setae on the

palm of gnathopod 2. J. L. Barnard’s Australian material shows marked

differences from the South African and Indian forms, especially as regards the

shape of the palm of gnathopod 2 and the setification of the telson. Further

sampling is necessary before it can be shown with certainty whether the various

morphs so far described are synonymous or represent different species.

Mallacoota subcarinata (Haswell, 1880)

Elasmopus subcarinatus: Stebbing, 1910a: 458.

Meaera subcarinata: K. H. Barnard, 1940: 460, fig. 26.

Mallacoota subcarinata: J. L. Barnard, 1972a: 247, figs 144, 145.

Records: SCD 262Q(2); LIZ 40C(3); MB 69R(1); 33°S/26°E/18-29 m (Stebbing

1910a).

Distribution: Mediterranean, Indo-Pacific.

Megaluropus agilis Hoek, 1889

Megaluropus agilis: Pillai, 1957: 50, fig. 10.

Records: MB 82H(1).

Diagnosis: Gnathopods simple; article 5 of gnathopod 2 dilated distally,

article 6 tapering distally; peduncle of uropod 3 short, rami greatly flattened,

subequal, their margins irregularly setose.

Distribution: North Atlantic, Mediterranean, India, South Africa.

Melita machaera K. H. Barnard, 1955

Melita machaera K. H. Barnard, 1955: 90-92, fig. 45.

Records: SCD 338T(1); TRA 54F(1).

292 ANNALS OF THE SOUTH AFRICAN MUSEUM

Diagnosis: Lower margins of articles 4-6 of gnathopod | thickly fringed with

short setae; palm of gnathopod 2 male slightly oblique, crenulate, a marked

semicircular concavity at its centre, defining angle bearing two teeth and a

pocket on its inner surface into which the tip of the dactyl closes; pleon seg-

ments 3 and 4 each with a small medio-dorsal tooth, segment 5 with a pair of

latero-dorsal teeth; uropod 3 as long as pleon segments 2-6 together.

Distribution: Endemic to south coast of South Africa.

Melita orgasmos K. H. Barnard, 1940

Melita orgasmos K. H. Barnard, 1940: 454. Sivaprakasam, 1966: 114, fig. 12 k—m.

Records: SCD 104Y(1); KNY 171E; LLL 6D; K 8N;; Still Bay, Port Elizabeth

(K. H. Barnard 1940).

Distribution: India, southern Africa.

Melita zeylanica Stebbing, 1904

Melita zeylanica: J. L. Barnard, 1972a: 235, figs 139-141.

Records: SUN 8K(8); BMR 7J(1), BMR 23T(1), BMR 25H(1), BMR 26S(8);

HAM 3R(3), HAM 9B(A), HAM 11G(P); HAV 3G, HAV 5D(1), HAV 7K(P),

HAV 8D(C), HAV 9B, HAV 10E(C), HAV 18J(A); KNY 50D, KNY 112A,

KNY 179C, KNY 291E; GBR 24G(2), GBR 46B(1); BRE 30Z(2), BRE 31E(3),

BRE 43F(5), BRE 44V(6), BRE 52N(2), BRE 71L(3), BRE 128A(1), BRE

135A(20), BRE 135C(1), BRE 144B(1), BRE 146A(4); STJ 7R, STJ 15K(CO),

STJ 16J(C), STJ 26F(1), STJ 27H(C), STI 29G, STJ 31U(A), STJ 32C; Little

Brak River, Keurbooms River, Wilderness lagoon, George, Port Elizabeth,

East London (K. H. Barnard 1940).

Distribution: Indo-Pacific (a brack-water species).

Parelasmopus ? suluensis (Dana, 1853)

Parelasmopus suluensis: Stebbing, 1888: 1029, pl. 100.

Records: QQ 4Y, S 54F.

Diagnosis: None of coxae 1-4 with antero-ventral teeth; palm of gnathopod 2

male oblique; article 2 of pereiopods 3-5 not grossly serrate, articles 3-5 not

very setose; pereon segment 7 and pleon segments 1-4 each with a pair of strong

dorsal teeth; third pleonal epimeron strongly serrate ventrally.

Distribution: ? Indo-Pacific.

Remarks: Considerable controversy surrounds the identity of this form (for

discussion see J. L. Barnard, 1972a: 253). The specimens described above con-

form with those of Stebbing (1888) but Stebbing’s material differs from that of

Dana (1853) in the presence of dorsal teeth on pereon segment 7 and in the

THE AMPHIPODA OF SOUTHERN AFRICA 293

weaker serrations on article 2 of pereiopods 3—5. Authors subsequent to Stebbing

have further confused the situation by synonymizing his P. suluensis with P.

setiger Chevreux, a move which it now appears was unjustified. As the situation

stands at present, definitive identifications cannot be made until the existing

forms are reviewed by someone who has access to the original types. In the

meantime I feel I cannot go further than to associate my material with the well-

known description of Stebbing (1888).

Family Haustoriidae

Bathyporeia sp.

Bathyporeia gracilis: K. H. Barnard, 1951: 704.

(non) Bathyporeia gracilis Sars 1891: 132-133, pl. 45 (4).

Records: SCD 244C(22), SCD 286H(2), SCD 329X(1), SCD 338M(1), SCD

349Q(1), SCD 370Q(1), SCD 376F(3), SCD 384R(1).

Distribution: Endemic, south coast of South Africa to South West Africa.

Remarks: This material was sent to Dr Wim Vader for use in a study on the

status of Bathyporeia gracilis Sars, but was found by him to be an undescribed

species (Vader 1970) which he is in the process of describing.

Cunicus n. gen.

Diagnosis: Antenna 1 geniculate between articles 2 and 3; antenna 2 geniculate

between articles 3 and 4 and 4 and 5; mandibular palp not arising from basal

process, molar represented by a spinose process; outer plate of maxilla 2 not

enlarged; maxillipedal palp 4-articulate; coxae 1 and 2 rounded-quadrate, of

moderate size; gnathopods subchelate, article 5 much longer than 6; pereiopods

with dactyls; uropods | and 2 represented by rounded setose lobes; uropod 3

biramous, rami subequal, the outer bi-articulate; telson cleft.

Type species: Cunicus profundus n. sp.

Relationships: This new genus lies closest to Urothoe but is distinguished from

it by the reduced first and second uropods (these are biramous in Urothoe), the

geniculate antennae and the unusual mandibular molar. :

Cunicus profundus n. sp.

Figs 8, 9

Description of male (3,5 mm): Body not more than three times as long as wide;

head as long as first two pereon segments, eyes absent; antennae short, subequal,

held in life in a folded position along side of head; articles 1 and 2 of antenna 1

equal, setose dorsally (Fig. 8B), article 3 somewhat shorter than 2, joint

294 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 8. Cunicus profundus n. gen., n. sp.

Male, 3,5 mm: A—lateral view; B—antenna 1; C—antenna 2; D—maxilla 1;

E—maxilla 2; F—maxilliped; G—mandible.

THE AMPHIPODA OF SOUTHERN AFRICA

Fig. 9. Cunicus profundus n. gen., n. sp.

Male, 3,5 mm: A—gnathopod 1; B—palm of gnathopod 1 enlarged; C—gnatho-

pod 2; D—palm of gnathopod 2 enlarged; E—pereiopod 3; F—ventral view of

urosome; G—telson.

295

296 ANNALS OF THE SOUTH AFRICAN MUSEUM

between between articles 2 and 3 geniculate, flagellum equal to accessory

flagellum, both 4-articulate; antenna 2 geniculate between articles 3 and 4 and

4 and 5, peduncle setose and spinose (Fig. 8C), flagellum 5—articulate; mandible

(Fig. 8G) with 3-articulate palp, incisor tridentate, lacinia mobilis bifid, spine

row of six serrate spines, molar represented by a spiniform process; palp of

maxilla 1 bi-articulate (Fig. 8D), tipped by three plumose setae, outer plate

bearing 12 strong spines, inner plate with two short setae; plates of maxilla 2

(Fig. 8E) terminating in a row of alternating simple and plumose setae; maxilli-

ped (Fig. 8F) with 4-articulate palp, outer plate with five distal plumose setae

and a medial row of alternating spines and setae, inner plate terminating in a

group of spines surrounded by short plumose setae.

Coxae rounded, 1-4 successively larger, distally setose, coxa 4 extending to

tip of article 2 of pereiopod 2; gnathopods 1 and 2 (Fig. 9A—D) subchelate,

palms minutely pectinate, transverse, defined by large spines; pereiopods 1 and

2 each with article 5 posteriorly spinose, dactyl basally surrounded by a ring of

spines; pereiopod 3 (Fig. 9E) greatly widened and strongly spinose, article 2

fringed by long plumose setae, article 4 with a fascicle of plumose setae arising

from its inner margin, article 5 twice as wide as long and with three transverse

rows of short strong spines and a marginal row of plumose setae on posterior

lobe, article 6 greatly produced postero-distally into a spinose and setose lobe

overhanging the simple cultriform dactyl; pereiopods 4 and 5 with plumose

setae on posterior margin of article 2 and a ring of strong spines around base of

dactyl.

Pleon somewhat narrower than pereon, pleonal epimera postero-distally

rounded, the second with a fascicle of plumose setae on its external surface;

urosome very short (Fig. 9F), uropods 1 and 2 represented only by rounded

setose lobes; uropod 3 of normal structure, peduncle quadrate, outer ramus

bi-articulate, slightly longer than the inner; telson cleft, extending to centre of

rami of uropod 3, apex of each lobe with three small setae.

Female: Ovigerous at 3,5 mm, bearing 3-5 large eggs, otherwise similar to male.

Holotype: SAM A13212, male, 3,5 mm.

Type Locality: SCD 384P, 33°52’'S/25°38’E, 15 December 1962, depth 7 m,

substrate grey sand.

Material: SCD 83B(1), SCD 128 U(1), SCD 225F(2), SCD 285P(1), SCD 286F(1),

SCD 350M(1), SCD 370P(1), SCD 384P(1).

Remarks: The body form of this species and the fact that it has been recovered

from depths of up to 0,6 m beneath the sand surface bear witness to its extra-

ordinary degree of adaptation to the fossorial mode of life. Morphological

adaptations include the broadly truncated body shape, wide pereiopods, large

coxae, reduced urosome and the loss of eyes. Other features also considered

advanced include the minimal degree of sexual dimorphism and small number

THE AMPHIPODA OF SOUTHERN AFRICA 297

of eggs carried by the female (3-5). These factors taken together indicate that

Cunicus is probably the most advanced burrowing amphipod known, certainly

the tendency for reduction of urosome and broadening of pereon are taken to

their extreme here.

Urothoe coxalis Griffiths, 1974

Urothoe coxalis Griffiths, 19746: 238, fig. 5.

Records: KNY 157B(9).

Distribution: Endemic, Durban to Knysna.

Urothoe elegans Bate, 1857

Urothoe elegans: Chevreux & Fage, 1925: 101, fig. 95.

Records: SCD 232F(1), SCD 392T(1); SST 5E(3), SST 16H(1), SST 19E(1),

SST 24J(4); 34°S/22°E/81 m (K. H. Barnard 1955).

Distribution: Atlantic and Indian Oceans.

Urothoe grimaldi Chevreux, 1895

Urothoe grimaldii: Chevreux & Fage, 1925: 99, fig. 93. K. H. Barnard, 1955: 84, fig. 41B.

Records: SCD 122D(1), SCD 135G(1), SCD 141M(7), SCD 146B(1), SCD

198L(95), SCD 202C(1), SCD 222N(1), SCD 225E(7), SCD 230C(1), SCD

232K(2), SCD 257N(14), SCD 2675(5), SCD 276L(10), SCD 278L(2), SCD

286D(10), SCD 315X(24), SCD 329V(8), SCD 348Z(4), SCD 368(T)1, SCD

370N(7), SCD 376G(13), SCD 381R(8); SST 1G(1), SST 61N(2), SST 73G(1);

MB 71H(6); 34°S/22°E/110 m, 34°S/22°E/81 m (K. H. Barnard 1955).

Distribution: Atlantic, Mediterranean, India, South Africa.

Urothoe platypoda n. sp.

Fig. 10

Description of male (2,5 mm): Head as long as three pereon segments; eyes

large, dark, subcircular; flagellum of antenna 1 5-articulate, accessory flagellum

of two equal segments; antenna 2 as long as body (Fig. 10A), terminal article

of peduncle bearing five large calceoli, flagellum slender, 26-articulate; palp of

maxilla | bi-articulate, outer plate bearing 10 strong spines; mandible (Fig. 10B)

with large quadrate molar and 3-articulate palp, article 1 of palp shorter than

2 or 3; maxilliped (Fig. 10C) with 4-articulate palp, article 2 broadly expanded

medio-distally, the expansion strongly setose, article 3 widening distally, outer

plate twice length of inner, bearing five strong spin-teeth medially, inner plate

terminating in five curved spines and a few short plumose setae.

Coxa 1 tapering to an acute apex, coxae 2-4 slightly produced backwards,

ESE

298 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 10. Urothoe platypoda n. sp.

Male, 2,5 mm: A—lateral aspect ; B— mandible; C—maxilliped; D—uropod 1; E—uropod 2; F—uropod 3; G—telson.

remaining coxae rounded; gnathopod | simple, article 5 expanded posteriorly,

bearing four spines postero-distally, article 6 small; gnathopod 2 subchelate,

palm strongly concave; pereiopods | and 2 strongly spinose posteriorly, dactyls

present; article 2 of pereiopod 3 unusually large, posterior margin scalloped,

postero-distal corner lobed to obscure part of article 3, article 4 with a single

postero-distal spine and a row of nine antero-distal spines, article 5 somewhat

broader than long, bearing two transverse rows of ten and nine spines anteriorly

and two rows of nine and six spines posteriorly (Fig. 10A), article 6 with three

THE AMPHIPODA OF SOUTHERN AFRICA 299

anterior rows of seven, five and three spines, a central group of five spines and

two single spines on posterior margin, dactyl broad, anteriorly serrate; article 2

of pereiopods 4 and 5 large, lobed postero-distally, articles 4 and 5 strikingly

broadened, 4 being as wide as 2.

Pleon segments very large, pleonal epimera postero-distally rounded, the

second bearing plumose setae on its outer surface; rami of uropod | (Fig. 10D)

equal, peduncle with two distal spines and each ramus with a single large spine

mid-dorsally; uropod 2 (Fig. 10E) much shorter than 1, peduncle with one

proximal and one distal spine, rami naked; uropod 3 (Fig. 10F) with quadrate

peduncle, rami foliacious, subequal, bearing long marginal plumose setae,

outer ramus with a short article 2; telson (Fig. 10G) 1,5 times as long as peduncle

of uropod 3, cleft to base, each lobe bearing a single spine and a minute seta at

its apex

Female (3 mm): Similar to the male except for the eyes, second antennae and

third uropods. The eyes are much smaller than those of the male, being composed

of about 12 well-spaced ommatidea, while antenna 2 is subequal to antenna 1,

lacks calceoli and has a 2-articulate flagellum. The third uropods do not exceed

the tip of the telson and have fewer, shorter marginal setae.

Holotype: SAM A13210, male, 2,5 mm.

Type locality: SCD 391G, 34°05’S/23°24’E, 8 December 1962, depth 11 m,

substrate yellow sand.

Relationships: The shape of pereiopods 3-5 serves to diagnose this species. The

second articles (especially that of pereiopod 3) are unusually large and the

enlarged flattened articles 4 and 5 of pereiopods 4 and 5 are unique in the genus.

Material: SCD 202D(6), SCD 230B(1), SCD 349N(1), SCD 391G(12).

Urothoe pinnata K. H. Barnard, 1955

Urothoe pinnata K. H. Barnard, 1955: 86, fig. 42.

Records: SCD 288G(1); SST 52N(5), SST 54B(2), SST 56P(3).

Distribution: Endemic, Natal to False Bay.

Urothoe pulchella (Costa, 1853)

Urothoe pulchella: Chevreux & Fage, 1925: 99, fig. 92. K. H. Barnard, 1955: 83, fig. 41A.

Records: SCD 74A(1), SCD 83A(2), SCD 104T(1), SCD 128S(2), SCD 135F(20),

SCD 148H(16), SCD 173W(1), SCD 202B(2), SCD 211V(3), SCD 285Q(6),

SCD 286E(1), SCD 287D(1), SCD 333R(4), SCD 345U(19), SCD 348Y(1),

SCD 349M(10), SCD 350L(1), SCD 383P(9), SCD 384N(45), SCD 391H(19);

SST 1E(2), SST 9B(2), SST 32C(1), SST 34H(3), SST 56U(3), SST 59K(3),

SST 60J(8), SST 61M(3), SST 62N(1), SST 63Q(1), SST 65E(9), SST 67M(4),

300 ANNALS OF THE SOUTH AFRICAN MUSEUM

SST 74D(1), SST 76M(1), SST 77D(7), SST 78D(5); MB 82C(2); SUN 5D(10);

HAV 7M, HAV 13L(2), HAV 17B(A); KNY 187B(3); GBR 12C(45), GBR

16J(7); STJ 14C(C), STJ 15L(P), STJ 16U(1), STJ 17C(1); 34°S/22°E/109 m

(K. H. Barnard 1916); 34°S/22°E/77 m, 34°S/22°E/110 m (K. H. Barnard 1955).

Distribution: Mediterranean, Atlantic, South Africa.

Urothoe tumorosa Griffiths, 1974

Urothoe tumorosa Griffiths, 1974b: 241, fig. 6.

Records: SST 54C(2).

Distribution: Endemic, Durban to Still Bay.

Family Ischyroceridae

Ischyrocerus anguipes Kroyer, 1838

Ischyrocerus anguipes: J, L. Barnard, 1954: 35, pls 32, 33; 19695: fig. 107B.

Records: SS 55L.

Distribution: Cosmopolitan in tropical and temperate seas.

Jassa falcata Montagu, 1808

Jassa falcata: Sexton & Reid, 1951: 30-47, pls 4-30. J. L. Barnard, 1969a: 115, figs 38, 39.

Records: MB 21E(1); KNY 166D, KNY 176C(A), KNY 179E(1); J 11K;

L 458A; E 234; ZZ 3N; T 13L; VV 2J; AR 1R(i).

Distribution: Cosmopolitan.

Parajassa chikoa n. sp.

Fig. 11

Description of male (2,5 mm): Head as long as three pereon segments, ocular

lobes moderately produced, distally rounded, eyes round; antenna | slightly

shorter than pereon, articles 2 and 3 subequal, 1,5 times as long as 1, flagellum

4-articulate, accessory flagellum not seen, presumed vestigial; mandible (Fig.

11B) with large 3-articulate palp, article 1 shorter than 2, 2 equal to 3, 3 strongly

setose distally, cutting edge of five strong teeth, lacinia mobilis apically bifurcate,

spine row of two strong spines, molar large, triturative; outer lobes of lower

lip simple, neither notched nor excavate; maxilla 1 (Fig. 11D) bearing bi-articu-

late palp, distal article ending in five serrate blade-spines and three subterminal

plumose setae, outer plate bearing seven strong serrate spines, inner plate simple;

maxilliped (Fig. 11E) with 4-articulate palp, outer plate bearing seven medio-

distal spines, inner plate with an oblique row of plumose setae and three small

distal spines.

THE AMPHIPODA OF SOUTHERN AFRICA 301

Fig. 11. Parajassa chikoa n. sp.

Male, 2,5 mm: A—lateral aspect; B—mandible; C—lower lip; D—mazxilla 1;

E—maxilliped; F—gnathopod 1; G—gnathopod 2; H—uropod 3; I—telson.

302 ANNALS OF THE SOUTH AFRICAN MUSEUM

Coxae 1-4 rounded, subequal, 5 bilobed, 6 and 7 each half as long as 5;

gnathopod 1 (Fig. 11F) subchelate, palm pectinate, defined by a single small

spine, dactyl slightly exceeding palm; gnathopod 2 (Fig. 11G) powerfully sub-

chelate, article 2 distally lobed, the lobe extending proximally as an anterior

keel, 5 cup-shaped, embracing the very large article 6, palm oblique, subequal

to hind margin, bearing a flat-topped tooth near finger hinge, a conical tooth

proximal to it then a strong semicircular concavity preceding the defining tooth,

dactyl moderately thickened, bearing a row of small setae posteriorly and a

fascicle of long setae at its apex; article 4 of pereiopods | and 2 with antero-dista!

corner expanded, dactyl long, slender; article 2 of pereiopod 3 subcircular,

article 6 bearing a distal spine; pereiopods 4 and 5 somewhat larger than 3 but

of similar structure.

Pleon and urosome flexed beneath pereon; pleonal epimera rounded;

urosomal segments subequal in length, uropods 1-3 extending equally; peduncle

of uropod | with five dorsal spines, inner ramus slightly exceeding outer, each

with a strong apical spine, outer ramus also with three dorsal spines; peduncle

of uropod 2 with three dorsal spines, each ramus with two dorsal and a terminal

spine, inner ramus slightly longer than outer; peduncle of uropod 3 (Fig. 11H)

twice length of rami, inner ramus tapering evenly, apex bearing a single minute

seta, outer ramus terminally upturned to form a pair of non-articulate hooks

(one specimen showed three hooks on the one uropod 3 and two on the other);

telson triangular, entire, smooth.

Holotype: SAM A13218, male, 2,5 mm.

Type locality: SCD 99M; 34°33’S/24°01’E, 21 July 1959, depth 130 m, substrate

rock.

Relationships: Of the four existing species in this genus Parajassa pelagica

Leach differs from P. chikoa n. sp. by virtue of its minute 2-articulate flagellum

of antenna 1, while P. angularis Shoemaker has distinctive setose first and second

uropods. The other two species, P. tristanensis (Stebbing) and P gorgoniana

(Schellenberg) are closely allied to P. chikoa n. sp. but have only a single tooth

on the palm of gnathopod 2 and smaller hooks on the outer ramus of uropod 3.

Material: SCD 99M(1), SCD 343Z(1). (Both 33).

Family Leucothoidae

Leucothoe ctenochir K. H. Barnard, 1925

Leucothoe ctenochir K. H. Barnard, 1925: 342, pl. 34, fig. 8.

Records: SCD 93N(2), SCD 159W(15), SCD 172S(2); MB 13M(1), MB 23J(1),

MB 28H(1), MB 54U(1).

Distribution: Endemic to east coast of South Africa.

THE AMPHIPODA OF SOUTHERN AFRICA 303

Leucothoe dolichoceras K. H. Barnard, 1916

Leucothoe dolichoceras K. H. Barnard, 1916: 157, pl. 26, fig. 14.

Records: SCD 181R(1); SST 11R(); 34°S/25°E/137 m, 32°S/28°E/93 m

(K. H. Barnard 1916).

Distribution: Endemic, Natal to west coast of South Africa.

Leucothoe richiardi Lessona, 1865

Leucothoe richiardii: K. H. Barnard, 1916: 150.

Leocothoe richiardi: Sivaprakasam, 19675: 385, fig. 2.

Records: SCD 24M(2), SCD 131X(1), SCD 159Q(1), SCD 181Q(1), SCD

204U(1), SCD 244F(1), SCD 310E(1), SCD 353H(1); SST 9D(1), SST 11Q(18);

SST 19D(1), SST 24F(2), SST 29B(2), SST 37V(1); MB 23F(4), MB 40F(1);

34°S/25°E/138 m, 33°S/28°E/121 m (K. H. Barnard 1916).

Distribution: Mediterranean, India, South Africa.

Leucothoe spinicarpa (Abildgaard, 1789)

Leucothoe spinicarpa: Sivaprakasam 19675: 384, fig. 1.

ineccorass “SCP 41R(2), SCD 95EC), SCD 115DQ), SCD 172TQ2), SCD

198X(1), SCD 253F(1), SCD 300U(1), SCD 302W(1), SCD 366M(3), SCD

366W(1); LIZ 37P(3), LIZ 40A(1); MB 23X(2), MB SOP(3), MB 54U(8), MB

69T(1); TRA 55W(1); LLL 6W; 32°S/28°E/174 m, 33°S/28°E/91 m (K. H.

Barnard 1916).

Distribution: Cosmopolitan.

Family Liljeborgiidae

Liljeborgia consanguinea Stebbing, 1888

Liljeborgia consanguinea Stebbing, 1888: 980, pl. 91.

Records: 35°S/20°E/565 m (Stebbing 1910a).

Diagnosis: Each of pleon segments 1—5 produced into a small but distinct mid-

dorsal tooth; coxae 1-3 each with a denticle at doth distal corners, coxa 4

serrate posteriorly; pleonal epimera 1 and 2 postero-distally produced into a

small point, that of third pleonal epimeron larger and upturned with a small.

sinus above; telson 60% cleft, a single spine at apex of each lobe; palm of

gnathopod 2 male smooth.

Distribution: Antarctica, southern Indian Ocean.

Liljeborgia dubia (Haswell, 1880)

Eusirus dubius Haswell, 1880: 331, pl. 30, fig. 3.

Records: SCD 24R(3), SCD 120C(1), SCD 216E(1), SCD 349S(1), SCD 353R(1),

SCD 366R(3); SST SH(1), SST 9G(1), SST 16R(5), SST 18X(1).

304 ANNALS OF THE SOUTH AFRICAN MUSEUM

Diagnosis: Pleon segments 1 and 2 each with five dorsal teeth, 3 with a minute

tooth between two rounded lobes, 4 and 5 each with a strong carinate dorsal

tooth; coxa 4 with two teeth on hind margin; postero-distal corner of third

pleonal epimeron acute, slightly upturned; telson cleft nearly to base, a long

spine in a notch at apex of each lobe; palm of gnathopod 2 male with an acute

distal tooth.

Distribution: Australia, New Zealand, South Africa.

Liljeborgia epistomata K. H. Barnard, 1932

Liljeborgia epistomata K. H. Barnard, 1932: 144, fig. 83; 1955: 89, fig. 44.

Records: SCD 110W(1), SCD 135M(1), SCD 141P(4), SCD 222J(1), SCD

DSA SED 376K).

Distribution: Endemic, Natal to Saldanha Bay.

Liljeborgia kinahani (Bate, 1862)

Liljeborgia kinahani: Chevreux & Fage, 1925: 157, fig. 157.

Records: SCD 262K(1), SCD 356K(1); MB 23H(2), MB 87F(3).

Diagnosis: Pleon segments | and 2 dorsally tridentate, 3 smooth, 4 and 5 each

with a single dorsal tooth; coxae 1-3 without distal teeth; coxa 4 not serrate

posteriorly; pleonal epimera | and 2 postero-distally produced into a minute

tooth, third pleonal epimeron with a small sinus above postero-distal tooth;

telson cleft nearly to base, lobes divergent, a long spine arising from a notch in

apex of each lobe; palm of gnathopod 2 male smooth.

Distribution: North Atlantic, South Africa.

Liljeborgia palmata n. sp.

Fig. 12

Description of male (6 mm): Head as long as first two pereon segments, rostrum

acute, slightly downturned, half length of article 1 of antenna 1, eyes absent;

antenna | as long as peduncle of antenna 2, article 1 considerably longer than

2 plus 3, flagellum 19-articulate, twice length of peduncle, accessory flagellum

10-articulate; flagellum of antenna 2 14-articulate, as long as terminal article

of peduncle; mandible (Fig. 12A) with 3-articulate palp, articles 1 and 2 sub-

equal, longer than 3, primary cutting edge with three large teeth and numerous

serrations, secondary cutting edge of five large teeth, spine row of eight spines,

molar redundant, represented by a few spines; inner plate of maxilla 1 tipped

by a single seta, outer plate bearing eight long pectinate spines, palp bi-articulate

with eight small spines lining inner edge and four small setae along outer margin;

plates of maxilla 2 subequal; inner plate of maxilliped bearing five terminal

setae, outer plate with a row of seven medial spines and seven submarginal

setae, palp 4-articulate.

THE AMPHIPODA OF SOUTHERN AFRICA 305

Fig. 12. Liljeborgia palmata n. sp.

Male, 6 mm: A—mandible; B—gnathopod 1; D—gnathopod 2; E—articles 2

and 3 of pereiopod 5; F—pleonal epimera 1-3; G—uropod 3; H—telson.

Male, 10 mm: C—gnathopod 2 (inner aspect).

306 ANNALS OF THE SOUTH AFRICAN MUSEUM

Coxae 1 and 2 each bearing two small postero-distal notches; gnathopod 1

(Fig. 12B) with a row of setae along anterior margin of article 2, articles 2 and

3 antero-distally lobed, 5 produced to protect 6 posteriorly, palm evenly convex,

bearing alternating long and short setae, defined by two spines, dactyl bearing

five proximal teeth; gnathopod 2 (Fig. 12D) larger than 1, article 2 bearing

two serrate anterior keels which are distally produced into moderate lobes,

article 3 with an antero-distal serrate lobe, article 6 tapering off from defining

angle, palm irregularly toothed and varying greatly with age (Fig. 12C, D),

dactyl with 4-6 proximal teeth, closing between defining spine and a spinose

ridge arising from inner face of article 6; pereiopods 1 and 2 slender; article 2

of pereiopods 3-5 widened, posteriorly serrate, the serrations most marked on

pereiopod 5, where they number 14.

Pleonal tooth formula 1:1:0:1:1, the teeth on the first two segments appressed

while those on segments 4 and 5 form carinae; postero-distal corner of pleonal

epimera 1 and 2 produced into a small tooth, third pleonal epimeron with a

semicircular concavity above postero-distal tooth and a second tooth bearing a

single small seta above this (Fig. 12F); uropods extending about equally,

peduncle of uropod | with a terminal spine, outer ramus marginally the shorter;

uropod 2 with two dorsal spines on peduncle, outer ramus slightly the shorter;

rami of uropod 3 (Fig. 12G) subequal, the outer naked but minutely pectinate

on upper margin, the inner with four dorsal and two ventral spines; telson

(Fig. 12H) 80% cleft, each lobe bearing a single strong seta in a subapical notch.

Female: Differing from the male only in the possession of brood plates and in

the structure of gnathopod 2, which is slightly larger than gnathopod 1, but of

the same structure.

Holotype: SAM A13221, male, 6 mm.

Type locality: SST 29G, 34°40’S/21°39’E, 21 June 1972, depth 80 m, substrate

coarse shelly sand.

Relationships: Males of this species can easily be distinguished by the unusual

shape of the palm of gnathopod 2 and by the serrate articles 2 and 3 of gnathopod

2. The female is similar to that of Liljeborgia hansoni Hurley, which is unfor-

tunately known only from the female. However, Hurley describes his species as

bearing red-brown eyes, whereas L. palmata n. sp. has no eyes.

Material: SST 16S(4), SST 29G(2).

Listriella sinuosa n. sp.

Fig. 13

Description of male (8 mm): Head only slightly longer than first pereon segment,

postantennal angle smoothly rounded, eyes small, round, black; antenna 1

shorter than peduncle of antenna 2, articles 1 and 2 subequal, 3 very short,

THE AMPHIPODA OF SOUTHERN AFRICA 307

Fig. 13. Listriella sinuosa n. sp.

Male, 8 mm: A-—lateral aspect; B—accessory flagellum; C—mandible; D—uropod 3; E—maxilla 1; F—telson.

flagellum 10-articulate, accessory flagellum (Fig. 13B) 4-articulate; antenna 2

as long as pereon, flagellum 10-articulate; mandible (Fig. 13C) with broad

3-articulate palp, articles 2 and 3 bearing rows of strong setae distally, incisor

strongly chitinized, primary cutting edge with seven strong teeth, spine row of

four spines, molar degenerate, represented by three setae; maxilla 1 (Fig. 13E)

with two apical setae on inner plate and eight strong pectinate spines on outer

plate, palp bi-articulate, tipped by three short spines and a row of short setae;

308 ANNALS OF THE SOUTH AFRICAN MUSEUM

inner plate of maxilliped with three setae along inner margin and a spine and

two setae at apex, outer plate with seven graduated spines along medial margin

and three setae distally, palp 4-articulate.

Coxa 1 produced anteriorly, coxae 2 and 3 rounded, 4 nearly twice as

long as 3, excavate posteriorly; gnathopod 1 much smaller than 2, palm evenly

convex, lined with alternating long and short setae; gnathopod 2 very large,

article 2 with two anterior keels, palm very long, bearing three small spines on

a rounded convexity near finger hinge, otherwise bearing scattered short setae;

pereiopods | and 2 slender; article 2 of pereiopods 3-5 widened; pereiopods 4

and 5 very elongate, hind margins of article 2 feebly serrate.

Pleonal epimera all smoothly rounded postero-distally; uropod 1 with a

row of about five dorsal peduncular spines, rami equal, spinose dorsally and

apically; uropod 2 without peduncular spines, otherwise resembling uropod 1;

uropod 3 (Fig. 13D) extending well beyond 1 and 2, peduncle bearing three

dorsal and several ventral spines, rami subequal, the outer with a spiniform

second article; telson (Fig. 13F) 80% cleft, each lobe bearing four large spines,

a smaller spine and a minute seta across its truncated apex.

Holotype: SAM A13215, male, 8 mm, unique.

Type locality: SST 41P, 34°25’S/21°28’E, 21 June 1972, depth 50 m, substrate

green mud.

Relationships: All other species in this genus, with the exception of L. lindae

Griffiths, have a bi-articulate accessory flagellum. L. sinuosa can be distinguished

from L. lindae by virtue of its smoothly rounded pleonal epimera and strongly

setose telson, as well as by the structure of gnathopod 2.

Family Lysianassidae

Amaryllis macrophthalma Haswell, 1880

Amaryllis macrophthalma: J. L. Barnard, 1972a: 262-269, figs 156-158.

Records: SCD 172X(5), SCD 181L(1), SCD 198C(3), SCD 204W(3), SCD

227S8(2), SCD 244Q(3), SCD 266S(3), SCD 300V(2), SCD 321V(1), SCD

379K(29); SST 16V(7), SST 29M(1); LIZ 17G(1), LIZ 29N(1), LIZ 37M(2);

MB 16D(1), MB 20U(1), MB 40H(1), MB 50Y(1), MB 66N(1), MB 73L(3);

KNY 57C(C), KNY 171H; TRA 55X(1), TRA 58X(1); LLL 6G; E 231A;

KKN 43E; ‘Algoa Bay’, 33°S/27°E/112 m (K. H. Barnard 1916); 33°S/26°E/18—

29 m (Stebbing 1908a).

Distribution: Cosmopolitan in southern hemisphere.

Aristias symbiotica K. H. Barnard, 1916

Aristias symbiotica K. H. Barnard, 1916: 121.

Records: SCD 55C(14).

Distribution: Endemic, Mogambique to South West Africa.

THE AMPHIPODA OF SOUTHERN AFRICA 309

Cyphocaris faurei K. H. Barnard, 1916

Cyphocaris faurei K. H. Barnard, 1916: 117, pl. 26, fig. 4.

Records: 33°S/28°E/450-550 m (K. H. Barnard 1916).

Diagnosis: First pereon segment greatly enlarged and swollen anteriorly but

not projecting over head; coxa 4 anteriorly and distally strongly convex, poste-

rior margin concave on either side of a medial tooth; article 2 of pereiopod 3

produced posteriorly as a curved spiniform process which extends to tip of

article 5, both margins of the process entire; telson as long as urosome, 75%

cleft, apices entire, lacking spines.

Distribution: Cosmopolitan, bathypelagic.

Euonyx conicurus K. H. Barnard, 1955

Euonyx conicurus K. H. Barnard, 1955: 80, fig. 38.

Records: SCD 102G(1), SCD 160C(3).

Diagnosis: Eyes present; article | of antenna 1 prominently lobed distally;

pleon segment 4 depressed at base then raised posteriorly into a large forward-

directed triangular carina.

Distribution: Endemic to south coast of South Africa.

Hippomedon onconotus (Stebbing, 1908)

Tryphosa onconotus Stebbing, 1908: 65, pl. 35.

Records: SCD 104V(1), SCD 104W(1), SCD 106U(1), SCD 120D(2), SCD

202F(3), SCD 204V(1), SCD 211X(1), SCD 216N(22), SCD 225M(2), SCD

235U(5), SCD 343V(6), SCD 350N(1), SCD 391L(4), SCD 392S(5); SST 18Y(2),

SST 24R(1), SST 62J(3), SST 63M(2), SST 65H(1), SST 67P(1), SST 68Q(3),

SST 70P(1), SST 74E(1).

Distribution: Endemic to South Africa.

Ichnopus taurus Costa, 1853

Ichnopus macrobetomma Stebbing, 1917: 38, pl. 96 A.

Ichnopus taurus: Chevreux & Fage, 1925: 48, fig. 30.

Records: 33°S/27°E/91 m (K. H. Barnard 1916); 33°S/28°E/47 m (Stebbing 1917).

Diagnosis: Eyes very large; antennae slender; upper lip slightly produced in

front of epistome; mandibular molar laminate, palp attached level with molar;

gnathopod 1 simple, slender, dactyl strongly spinose posteriorly; dactyl of

gnathopod 2 minute; gills plated on both sides; pereiopod 5 much longer than 4;

inner ramus of uropod 2 constricted; telson deeply cleft.

Distribution: Mediterranean, North Atlantic, South Africa.

310 ANNALS OF THE SOUTH AFRICAN MUSEUM

Lepidepecreum twalae n. sp.

Fig. 14

Description of female (2,5 mm): Head dorsally shorter than first pereon segment,

produced below antenna | into a large apically rounded lobe, eyes present but

almost colourless in preserved animal; article 1 of antenna | (Fig. 14B) laterally

flattened but not dorsally carinate, almost as tall as head, articles 2 and 3 much

smaller, flagellum 5-articulate, shorter than peduncle, accessory flagellum

small, bi-articulate; antenna 2 (Fig. 14C) not much longer than 1, article 3

elongate, flagellum 3-articulate, about as long as terminal article of peduncle;

mandible (Fig. 14D) with smooth incisor, spine row of three spines, molar

setulose, of moderate size, palp 3-articulate, attached proximal to molar, article

2 very elongate, article 3 finely setulose throughout and bearing three lateral

and three terminal setae; maxilla 1 (Fig. 14E) with bi-articulate palp, article 2

terminating in 6 blade spines and a small seta, outer plate bearing ten powerful

serrate spines, inner plate with two setae at its tip; plates of maxilla 2 (Fig. 14F)

subequal; maxilliped (Fig. 14G) with 4-articulate palp, outer plate bearing 12

nodular spines along medial margin and an oblique row of five larger spines

submarginally, inner plate with two distal spines and six medial plumose

setae.

Pereon dorsally smooth, coxae 1-4 all visible, elongate and distally touching

their partners on the opposite side, coxa 4 excavate posteriorly, 5—7 subcircular;

gnathopod | (Fig. 141) subchelate, palm transverse, defined by two large spines,

dactyl with accessory tooth; gnathopod 2 (Fig. 14H) chelate, article 5 slightly

longer than 6, both distally setulose; article 2 of pereiopods 1 and 2 slender;

article 2 of pereiopods 3-5 rotund, article 4 produced into a postero-distal

lobe.

Pleon segments 1-3 dorsally smooth, pleonal epimera postero-distally

rounded; pleon segment 4 bearing a large dorsal triangular carina; uropods 1-3

all projecting equally; uropod 3 (Fig. 14K) with equal rami, inner ramus with

three large spines on dorsal margin, outer ramus with two spines at apex of

article 1, article 2 triangular, tapering evenly; telson (Fig. 14L) twice as long

as wide, 70% cleft, each lobe tipped by a small spine and with a short mid-

dorsal plumose seta.

Holotype: SAM A13220, female, 2,5 mm, unique.

Type locality: SCD 343W, 36°39’S/23°41’E, 11 February 1962, depth 121 m,

substrate shelly sand.

Relationships: This species can be distinguished from most others in the genus

by the lack of carinae on the pereon and on article 1 of antenna 1. These features

are also absent in Lepidepecreum typhlops Bonnier and L. cingulatum K. H.

Barnard, but of these the former lacks eyes and the latter does not have a carina

on the fourth pleon segment.

THE AMPHIPODA OF SOUTHERN AFRICA 311

Fig. 14. Lepidepecreum twalae n. sp.

Female, 2,5 mm: A—lateral aspect; B— antenna 1; C—antenna 2; D—mandible;

E—maxilla 1; F—maxilla 2; G—maxilliped; H—gnathopod 2; I—gnathopod 1;

J—enlargement of palm of gnathopod 1; K—uropod 3; L—telson.

S12 ANNALS OF THE SOUTH AFRICAN MUSEUM

Lysianassa ceratina (Walker, 1889)

Lysianassa cubensis: K. H. Barnard, 1916: 120.

Lysianassa ceratina: Chevreux & Fage, 1925: 42, fig. 23.

Records: SCD 160D(9), SCD 173V(2), SCD 211Y(3), SCD 230E(1), SCD

235Z(1), SCD 310D(28), SCD 373N(15), SCD 374J(8); SST 47C(3), SST 47D(3);

LIZ 25M(1), LIZ 32Z(1); MB 40G(1), MB 54Y(1), MB 57A(4), MB 73N(4);

KNY 166K; BRE 144C(1); L 476; X 22D; Y 12H; E 231; ZZ 30: dia

KN 2E; SS 55A; 33°S/27°E/120 m, 33°S/28°E/78 m (K. H. Barnard 1916);

Port Elizabeth, East London (K. H. Barnard 1940).

Distribution: Mediterranean, Atlantic, Indian Ocean.

Lysianassa variegata (Stimpson, 1855)

Lysianassa variegata: Stebbing, 1888: 682, pl. 23.

Records: SCD 10U(2), SCD 148L(3), SCD 179M(1), SCD 181¥(1), SCD

189D(1), SCD 204T(1), SCD 262L(1), SCD 312S(2), SCD 338D(20), SCD

370M(14), SCD 376M(8), SCD 388E(1); SST 16W(1), SST 29N(1); LLL 6F;

33°S/26°E/18-29 m (Stebbing 1910a).

Distribution: Africa south of the equator.

Orchomene plicata (Schellenberg, 1925)

Orchomenopsis chilensis Schellenberg 1925: 119, fig. 3. K. H. Barnard 1925: 330.

Orchomenella plicata: K. H. Barnard, 1940: 440.

Records: SCD 20K(A), SCD 24G(3), SCD 287A(88); SST 65J(8); MB 28E(2),

MB 61T(1), MB 73P(3); KNY 57C; Plettenberg Bay (K. H. Barnard 1940—as

Microlysias xenoceras).

Distribution: Cosmopolitan.

Phoxostoma algoense K. H. Barnard, 1925

Phoxostoma algoense K. H. Barnard, 1925: 323, pl. 34, fig. 2.

Records: Algoa Bay, 65 m (K. H. Barnard 1925).

Diagnosis: Eyes large, meeting on top of head; body lacking carinae but with

scattered setules; mouthparts forming a conical bundle; mandible slender,

molar obsolete, palp attached proximal to molar; palp of maxilliped exceeding

outer plate, article 4 small; gnathopod | simple, article 6 longer than 5; gnatho-

pod 2 minutely chelate; third pleonal epimeron quadrate, telson deeply insinuate.

Distribution; Endemic to South Africa.

THE AMPHIPODA OF SOUTHERN AFRICA 313

Socarnopsis crenulata Chevreux, 1910

Socarnopsis crenulata: K. H. Barnard, 1916: 124. Chevreux & Fage, 1925: 49, figs 31, 32.

Records: SCD 81E; SST 17Z(1), SST 29P(8), SST 32K(4), SST 34M(2), SST

45H(1); 33°S/27°E/120 m (K. H. Barnard 1916).

Diagnosis: Eyes large; upper lip and epistome together produced forwards into

a lobe; mandibular molar ridged, palp attached proximal to molar, its second

article very elongate; gnathopod | simple; dactyl of gnathopod 2 minute; gills

plaited on both sides; outer ramus of uropod 3 with a minute second article;

telson 80% cleft.

Distribution: Mediterranean, Atlantic, South Africa.

Stomacontion prionoplax Monod, 1937

Fig. 15

Stomacontion prionoplax Monod, 1937: 6, figs. 1-6.

Records: SCD 179N(1), SCD 244T(1).

Diagnosis: Coxa 1 rectangular, projecting over the side of the head and partly

obscuring the eye; gnathopod 1 simple, article 5 as wide as long, dactyl powerful;

articles 2 and 3 of pereiopod 5 strongly lobed anteriorly; pleon segment 4

bearing a triangular dorsal carina; uropod 3 lacking rami or rami vestigial.

Distribution: This is the first record of this species from southern Africa. It was

previously known only from the Suez Canal.

Remarks: The present material so closely resembles Monod’s that I have no

hesitation in equating the two. Of particular note is the identical structure of the

mouthparts, particularly the unusual outer plate of the maxilliped, and of

pereiopod 5. The specimen figured here is much larger (5 mm) than Monod’s

and has smaller eyes. My other specimen, however, is of comparable size to

Monod’s and has similar sized eyes. The only significant feature distinguishing

the South African form is the absence of the obscure vestigial ramus of uropod

3 described by Monod. The apparent variability of this feature, in addition to

other intergrading features, renders the maintenance of any distinction between

the genera Stomacontion and Acontiostoma superfluous, at least as they are.

defined at present, and the two genera should be united.

Trischizostoma remipes Stebbing, 1908

Trischizostoma remipes Stebbing, 1908a: 61, pl. 34. K. H. Barnard, 1925: 321.

Records: SCD 159P(10); SST 47A(2); 33°S/28°E/86 m, 34°S/23°E/58 m (Stebbing

1908a); “Cape Point to East London’ (K. H. Barnard 1916).

Distribution: Endemic, Natal to False Bay.

314 ANNALS OF THE SOUTH AFRICAN MUSEUM

SHI

Fig. 15. Stomacontion prionoplax Monod, 1937

Male, 5 mm: A-—lateral aspect; B—uropod 3; C—telson; D—mandible;

E—maxilla 1; F—maxilliped ;G—gnathopod 1 ; H—gnathopod 2; I—pereiopod 5.

THE AMPHIPODA OF SOUTHERN AFRICA 315

Trischizostoma serratum K. H. Barnard, 1925

Trischizostoma serratum K. H. Barnard, 1925: 320, pl. 34, fig. 1.

Records: SCD 366L(1).

Distribution: Endemic, Natal to False Bay.

Tryphosella africana K. H. Barnard, 1955

Tryphosella africana K. H. Barnard, 1955: 81.

Records: SCD 135E, SCD 198B(38), SCD 227T(1), SCD 236B(1), SCD 273F(3),

SCD 338S(1), SCD 345W(9), SCD 348Z(11), SCD 349P(1), SCD 356U(2),

SCD 376D(3), SCD 381U(1); SST 16X(1), SST 24S(2), SST 34L(1), SST 45G(2),

SST 65C(5).

Diagnosis: Lateral lobes of head moderately acute, eyes absent; article 4 of

antenna | male broadly oval, width more than half length; article 6 of gnathopod

1 ovoid, palm oblique, equal to hind margin; third pleonal epimeron greatly

produced postero-distally into an acute upturned tooth; pleon segment 4 with

a rounded dorsal hump; telson tapering evenly, bearing two pairs of dorsal

spines and a spine and seta at apex of each lobe.

Distribution: Endemic to South Africa.

Tryphosella normalis K. H. Barnard, 1955

Tryphosella normalis K. H. Barnard, 1955: 80, fig. 39.

Records: SCD 110X(4), SCD 173Z(2), SCD 192Z(7); SST 5C(1), SST 18Z(4),

SST 24T(1), SST 32L(1), SST 34D(3), SST 52R(3), SST 54F(1), SST 60L(2),

SST 74F(1), SST 78G(2).

Distribution: Endemic, Natal to South West Africa.

Uristes sulcus n. sp.

Fig. 16

Description of female (3 mm): Head dorsally shorter than first two pereon

segments, lateral lobes subacute, extending half length of article 1 of antenna 1;

eyes obscure, composed of about seven scattered ommatidea; article 1 of

antenna | large, longer than 2 and 3 together, flagellum 6-articulate, the first

segment much the largest, accessory flagellum 3-articulate; antenna 2 exceeding

1, flagellum 9-articulate; mandible (Fig. 16B) with smooth heavily chitinized

incisor, spine row of three spines, molar large, oval, ridged, palp 3-articulate,

attached level with distal margin of molar; maxilla 1 (Fig. 16C) with two plumose

setae at apex of inner plate, nine strong serrate spines at tip of outer plate, palp

bi-articulate, terminating in nine small spine-teeth; palp of maxilliped (Fig. 16D)

316 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 16. Uristes sulcus n. sp.

Female, 3 mm. A—lateral aspect; B—mandible; C—maxilla 1; D—maxilliped;

E—gnathopod 1; F—gnathopod 2. Male, 4 mm: G—uropod 3; H—telson.

THE AMPHIPODA OF SOUTHERN AFRICA 317

4-articulate, article 2 slightly longer than 1, article 4 small, outer plate with a

row of nine spines along medial edge, inner plate bearing two plumose setae

and two terminal spines.

Coxa 1 tapering evenly, about 80% length of coxa 2 which is rectangular

and expanded somewhat distally; gnathopod | (Fig. 16E) subchelate, article 6

slightly longer than 5, palm oblique and subequal to hind margin, defined by

two strong spines, dactyl without strong accessory tooth; gnathopod 2 (Fig. 16F)

with setulose distal articles, 6 shorter than 5, palm transverse, dactyl small,

partially concealed by setae; pereiopods | and 2 slender; article 2 of pereiopods

3-5 large, oval, bearing successively more obvious horizontal ridges.

Pleon segments 1-3 large, the integument bearing numerous longitudinal

furrows and ridges, these being particularly obvious on the dorsal surface of

pleon segment 3; pleonal epimera smoothly rounded; postero-dorsal section of

pleon segment 3 produced as a hood-like lobe arching over proximal portion of

pleon segment 4; pleon segment 4 with a small dorsal carina distally; outer

ramus of uropod | slightly the longer, bearing three dorsal spines, inner ramus

with two dorsal spines; outer ramus of uropod 2 with two dorsal spines, inner

ramus not constructed, bearing a single medio-dorsal spine; peduncle of uropod

3 with a single dorsal spine and three ventral spines, rami naked, lanceolate,

inner equal to article 1 of the outer; telson 80% cleft, a dorsal seta, two dorsal

spines and a terminal spine to each lobe.

Male: The male of this species differs from the female by virtue of its longer

second antennae, which reach half the length of the pereon, by the slightly

larger carina on pleon segment 4 and by the setose third uropods (Fig. 16G).

Holotype: SAM A13224, male, 4 mm.

Type locality: SCD 230D, 34°04’S/23°26’E, 29 November 1960, depth 43 m,

substrate yellow sand.

Relationships: J. L. Barnard (1962), in his revision of the genus Uristes, has

drawn into it species previously assigned to some five other genera. The genus

now contains species in which the condition of gnathopod 1 ranges from simple

through to fully subchelate. Uristes sulcus n. sp. falls into that section, having

an oblique, well-defined palm of gnathopod 1. It can be distinguished from

others in that group by the presence of eyes and by the rounded pleonal epi-

meron, as well as by virtue of the unusual ridging on the pereiopods and pleon

from which its name has been derived.

Material: SCD 230D(2); SST 29L(1), SST 37X(2).

Uristes natalensis K. H. Barnard, 1916

Uristes natalensis K. H. Barnard, 1916: 126.

Records: 33°S/27°E/110 m (K. H. Barnard 1916).

Distribution: Endemic to south and east coasts of South Africa.

318 ANNALS OF THE SOUTH AFRICAN MUSEUM

Family Ochlesidae

Ochlesis lenticulosus K. H. Barnard, 1940

Ochlesis lenticulosus K. H. Barnard, 1940: 447, fig. 23.

Records: SCD 244G(1); SST 11V(4).

Distribution: Endemic, Natal to False Bay.

Ochlesis levetzowi Schellenberg, 1953

Ochlesis levetzowi Schellenberg, 1953: 115, fig. 4. J. L. Barnard, 19695; 372, fig. 134a.

Records: SCD 308G(5).

Distribution: Endemic, south coast of South Africa to South West Africa.

Family Oedicerotidae

Perioculodes longimanus (Bate & Westwood, 1868)

Perioculodes longimanus: Chevreux & Fage, 1925: 162, figs 163, 164.

Records: SCD 115E(1), SCD 122H(1), SCD 128T(1), SCD 135C(1), SCD

198V(6), SCD 202G(1), SCD 211W(3), SCD 225K(2), SCD 227X(1), SCD

244L(2), SCD 257K(5), SCD 267K(1), SCD 273J(3), SCD 276N(1), SCD

278H(7), SCD 282T(2), SCD 285M(2), SCD 329Z(1), SCD 332T(2), SCD

338F(18), SCD 339S(1), SCD 343X(1), SCD 345X(1), SCD 368R(1), SCD

3763(3), SCD 379P(3), SCD 381T(1), SCD 383S(1), SCD 391M(5); SST 24K(4),

SST 29K(1), SST 32H(1), SST 34G(2), SST 40T(2), SST 45C(3), SST 52S(5),

SST 54F(1), SST 56R(1), SST 60K(1), SST 61L(2), SST 63P(1), SST 65F(7),

SST 67N(1), SST 73H(1), SST 76N(1), SST 78E(1), SST 81C(1).

Distribution: Mediterranean, Atlantic, Indian Ocean.

Westwoodilla manta n. sp.

Fig. 17

Description of male (6 mm): Head as long as first three pereon segments, rostrum

almost half length of head (Fig. 17A), rod-like in shape, not deflexed, its apex

smoothly rounded; eyes dorsally coalesced, situated at apex of rostrum; post-

antennal angle of head rounded; antenna | slightly exceeding peduncle of

antenna 2, flagellum 14-articulate; antenna 2 as long as pereon, flagellum

59-articulate; mandible (Fig. 17B) with 3-articulate palp, second article setose,

moderately curved, article 3 subequal to 2, article 1 short, incisor of mandible

not toothed but strongly chitinized, lacinia mobilis consisting of asmall flattened

plate, spine row of three spines, molar large, weakly ridged.

Coxa 1 marginally setose, distally expanded and produced forwards as a

THE AMPHIPODA OF SOUTHERN AFRICA 319

Fig. 17. Westwoodilla manta n. sp.

Male, 6 mm: A—head; B—mandible; C—gnathopod 1; D—gnathopod 2;

E—uropod 3; F—telson.

broad lobe; gnathopod | subchelate (Fig. 17C), article 5 moderately lobed, the

lobe not protecting article 6 posteriorly, palm oblique, bearing alternating long

and short setae, defined by a single small spine; coxa 2 rectangular; article 5

of gnathopod 2 (Fig. 17D) less strongly lobed than that of gnathopod 1, the lobe

projecting at right-angles, not protecting article 6, palm oblique, setose, defined.

by a small spine; coxae 3 and 4 oval, 4 not excavate posteriorly, coxa 5 bilobed;

pereiopods 1 and 2 slender, articles 4-6 strongly setose posteriorly, dactyl

powerful, longer than article 6; pereiopods 3-5 successively larger, 5 very

elongate.

Pleonal epimera smoothly rounded postero-distally; uropods 1-3 very

slender, projecting equally, in each case outer ramus fractionally the shorter;

telson (Fig. 17F) smoothly rounded, bearing four small setae distally.

Holotype: SAM A13223, male, 6 mm, unique.

320 ANNALS OF THE SOUTH AFRICAN MUSEUM

Type locality: SCD 24W, 34°07'S/23°23’E, 26 May 1958, depth 46 m, substrate

rock.

Relationships: The very elongate straight rostrum of this species is unusual.

Most other members of the genus show a deflexed rostrum seldom extending

to the tip of article 1 of antenna 1. Exceptions to this norm include Westwoodilla

acutifrons Gurjanova, W. longidactyla Carausu and W. rectirostris Chevreux.

However the rostrum of W. acutifrons, as the name suggests, terminates acutely

and the eyes are medial rather than terminal. In W. longidactyla the eyes occupy

the whole rostrum and the flagellum of antenna 1 does not exceed the length of

articles 2 plus 3 of the peduncle. W. rectirostris bears the closest resemblance

to W. manta n. sp. but differs from it by having longer antennae and the tip of

the rostrum produced into an acute-tipped upturned process.

Family Paramphithoidae

Epimeria cornigera (Fabricius, 1779)

Epimeria cornigera: Chevreux & Fage, 1925: 191, figs 198-200.

Records: 33°S/28°E/550 m (K. H. Barnard 1916).

Diagnosis: Pereon segment 7 (and sometimes 6) and pleon segments 1-3 each

bearing a pronounced dorsal carina flanked by a pair of subdorsal ridges;

pleon segment 4 bearing a strong dorsal carina terminating in an acute tooth;

coxae 1-5 all terminating acutely; third pleonal epimeron with an accessory

tooth on posterior margin above the acute postero-distal corner.

Distribution: North Atlantic, Mediterranean, South Africa.

Family Pardaliscidae

Nicippe tumida Bruzelius, 1859

Nicippe tumida: J. L. Barnard, 1959b: 39-40, figs 1, 2.

Records: SCD 95K(1).

Diagnosis: Article 2 of antenna 1 shorter than article 1; gnathopods slightly

subchelate, articles 5 and 6 stout, 5 with a large posterior lobe, shorter than 6,

palm undefined; article 2 of pereiopods 3-5 not inflated; pleon segment 4

bearing two contiguous small dorsal teeth; telson deeply cleft.

Distribution: Cosmopolitan.

Pardisynopia anacantha (K. H. Barnard, 1925)

Halice anacantha K. H. Barnard, 1925: 347, pl. 34, fig. 12.

Pardisynopia anacantha: J. L. Barnard, 19695: 400.

Records: SCD 343Y(4); SST 17Y(2).

Diagnosis: Article 2 of antenna 1 shorter than article 1; flagellum of antenna 1

THE AMPHIPODA OF SOUTHERN AFRICA 321

fully segmented, article 1 less than half length of peduncular article 1, accessory

flagellum 2-articulate, basal article as long as articles 1-3 of primary flagellum;

eyes absent; gnathopods simple; pereon and pleon without any dorsal teeth;

third pleonal epimeron postero-distally quadrate; telson twice as long as broad,

cleft nearly to base, lobes strongly dehiscent, their apices bifid.

Distribution: Endemic to South Africa.

Remarks: The above observations on the condition of antenna | confirm J. L.

Barnard’s provisional placing of this species in Pardisynopia (J. L. Barnard

19695).

Family Phoxocephalidae

Mandibulophoxus stimpsoni (Stebbing, 1908)

Pontharpinia stimpsoni Stebbing, 1908a: 75, pl. 11.

Mandibulophoxus stimpsoni: J. L. Barnard, 1957: 436-438, figs 3, 4.

Records: SCD 83C(11), SCD 94X(1), SCD 1048(4), SCD 120A(1), SCD

122C(16), SCD 128R(9), SCD 138B(18), SCD 141N(1), SCD 141S(2), SCD

148G(15), SCD 151F(1), SCD 173S(2), SCD 188B(12), SCD 194V(3), SCD

198W(1), SCD 199G(3), SCD 202A(1), SCD 204P(3), SCD 211S(6), SCD

216N(1), SCD 222K(1), SCD 225G(4), SCD 228A(3), SCD 320A(6), SCD

232H(2), SCD 235V(1), SCD 262P(1), SCD 267H(4), SCD 273G(1), SCD

285K(11), SCD 286G(1), SCD 287B(1), SCD 288H(1), SCD 300T(1), SCD

310J(10), SCD 315W(8), SCD 321S(1), SCD 326J(1), SCD 329T(8), SCD

338B(15), SCD 343S(3), SCD 345T(9), SCD 348X(15), SCD 349L(8), SCD

350K(14), SCD 353K(2), SCD 356Q(2), SCD 370L(5), SCD 376E(9), SCD

381N(9), SCD 383R(3), SCD 384Q(4), SCD 391G(14), SCD 392V(4); SST 1C(1),

SST 5D(3), SST 9A(3), SST 16C(2), SST 19C(4), SST 24G(10), SST 29A(52),

SST 32A(7), SST 34C(17), SST 52Q(2), SST 54A(2), SST 56Q(16), SST 60H(13),

SST 61K(4), SST 62L(3), SST 63N(1), SST 65A(7), SST 67L(4), SST 68P(8),

SST 70N(7), SST 73E(6), SST 74C(5), SST 76L(3), SST 77B(18), SST 78F(2):

MB 66U(4), MB 71G(3); 33°S/28°E/86 m, 33°S/26°E/18-29 m (Stebbing 1910a).

Distribution: West and South Africa.

Paraphoxus oculatus Sars, 1891

Paraphoxus oculatus: J. L. Barnard, 1960: 240-243, pls 27, 28.

Records: SST 81D(2).

Diagnosis: Rostrum tapering evenly in front of the eyes, apically rounded;

third pleonal epimeron not produced postero-distally, outer surface without a

setal row; article 6 of gnathopods 1 and 2 widened; telson cleft almost to base.

Distribution: Circumboreal.

522 ANNALS OF THE SOUTH AFRICAN MUSEUM

Platyischnopus herdmani Walker, 1904

Platyischnopus capensis K. H. Barnard, 1925: 338, pl. 34, figs 13, 14.

Platyischnopus herdmani: Rabindranath, 1971: 521, figs 1, 2.

Records: SCD 198J(2), SCD 211U(1), SCD 244D(3), SCD 338N(2), SCD

376H(1); SCD 381W(1); SST 29J(1), SST 45B(1), SST 56T(1), SST 59J(1),

SST 77C(1).

Distribution: India, South Africa.

Remarks: The genus Platyischnopus has been moved from Haustoriidae to

Phoxocephalidae as proposed by Bousfield (1970).

Family Podoceridae

Laetmatophilus purus Stebbing, 1888

Laetmatophilus purus Stebbing, 1888: 1198, pl. 132.

Records: SCD 95H(1), SCD 127P(2), SCD 198K(1), SCD 204S(2), SCD 211T(2),

SCD 216B(1); SST 32J(1); Q 7K; 34°S/25°E/138 m (K. H. Barnard 1916).

Distribution: Endemic, Mogambique to South West Africa.

Podocerus africanus K. H. Barnard, 1916

Podocerus africanus K. H. Barnard, 1916: 278, pl. 28, figs 24, 25; 1937: 176, fig. 19.

Records: LIZ 40D(8); J 11D; L 503; SS 55N.

Distribution: Arabia, Natal to South West Africa.

Podocerus brasiliensis (Dana, 1853)

Podocerus brasiliensis: J. L. Barnard, 1971: 117, figs 58-60.

Records: L 484, L 485.

Distribution: Cosmopolitan in tropical and temperate seas.

Podocerus cristatus (Thompson, 1879)

Podocerus cristatus: J. L. Barnard, 1962: 67, fig. 31.

Records: SCD 95B(2), SCD 102F(3), SCD 122G(1), SCD 302R(2), SCD

308H(4), SCD 392W(2); KNY 166L.

Distribution: Cosmopolitan in tropical and warm temperate seas.

Podocerus hystrix Stebbing, 1910

Podocerus hystrix Stebbing, 1910b: 622, pl. 58.

Records: SST 11P(16).

Diagnosis: Head with very short triangular rostrum and a large forward-directed

THE AMPHIPODA OF SOUTHERN AFRICA 323

medio-dorsal process; pereon segment | with two medio-dorsal carinae; remain-

ing pereon segments and pleon segments | and 2 bearing successively larger

backward-directed processes each flanked by a pair of smaller lateral processes;

lateral margins of pereon segments produced over the coxae; coxa | acutely

produced forwards, 2-4 acutely produced ventrally; palm of gnathopod 2

defined by two spines, that of male with a small tooth near finger hinge.

Distribution: This is the first record of this species from Africa. It was previously

known only from Australia.

Podocerus inconspicuus (Stebbing, 1888)

Podocerus palinuri K. H. Barnard, 1916: 277, pl. 28, fig. 23.

Podocerus inconspicuus: Nagata, 1965: 322, fig. 43.

Records: SCD 24H(1), SCD 55J(1), SCD 135A(2), SCD 146C(2), SCD 172W(2),

SCD 208J(1), SCD 249T(1), SCD 283G(2), SCD 302S(3), SCD 345V(5), SCD

353E(6), SCD 356R(1); SST 24H(1), SST 29H(2); MB 16E(2), MB 57C(1),

MB 87E(3); SS 55M; 34°S/22°E/215 m, 34°S/23°E/84 m (K. H. Barnard 1916).

Distribution: Indo-Pacific, west coast of South Africa.

Podocerus multispinis K. H. Barnard, 1925

Podocerus multispinis K. H. Barnard, 1925: 367, pl. 34, fig. 18.

Records: SCD 208F(2), SCD 353D(1); 34°S/25°E/128 m (K. H. Barnard 1916).

Distribution: Endemic, Natal to Saldanha Bay.

Family Sebidae

Seba saundersi Stebbing, 1875

Paravalettia chelata K. H. Barnard, 1916: 112, pl. 26, figs 2, 3.

Seba saundersii: K. H. Barnard, 1957: 7, fig. 4.

Records: 33°S/28°E/? (K. H. Barnard 1957).

Diagnosis: Eyes absent; gnathopod | subchelate in juvenile, becoming fully

chelate in adult, article 6 with a notch on posterior margin bearing 4-5 plumose

setae, dactyl closely fitting palm; gnathopod 2 smaller than 1, slender, chelate;

article 4 of pereiopods 3 and 4 acutely produced half length of article 5, that of

pereiopod 5 strongly expanded; telson entire, triangular, apically bluntly

rounded.

Distribution: Cosmopolitan.

Family Stegocephalidae

Parandania boecki (Stebbing, 1888)

Parandania boecki: J. L. Barnard, 1961: 57, fig. 27.

Records: 33°S/28°E/900 m (K. H. Barnard 1916).

324 ANNALS OF THE SOUTH AFRICAN MUSEUM

Diagnosis: Accessory flagellum of antenna | uni-articulate, almost as long as

article 1 of primary flagellum, which exceeds half length of peduncle; mandibular

incisor untoothed; palp of maxilla 1 uni-articulate; article 2 of pereiopod 3

slender, that of pereiopods 4 and 5 broad; telson oval, entire.

Distribution: Cosmopolitan, pelagic.

Stegocephaloides australis K. H. Barnard, 1916

Stegocephaloides australis K. H. Barnard, 1916: 129, pl. 28, fig. 29.

Records: SCD 181K(1); SST 1F(1), SST 9E(1), SST 11U(11), SST 19F(1),

SST 24P(1).

Diagnosis: Head almost hidden under the tumid first pereon segment; eyes

absent; coxae forming a continuous shield, 4 deeper than its pereon segment,

the posterior and distal margins differentiated by a rounded angle; article 2

of pereiopod 5 distally produced into an evenly rounded process reaching to

apex of article 4; hind margin of article 2 of pereiopod 5 feebly serrate; third

pleonal epimeron postero-distally rounded.

Distribution: Endemic to South Africa.

Family Stenothoidae

Parametopa grandimana n. sp.

Fig. 18

Description of male (3 mm): Head slightly shorter than two pereon segments,

eyes of moderate size, colourless; antenna 1 exceeding length of body, articles

1 and 2 subequal, each more than twice length of article 3, flagellum of 25

elongate articles, twice as long as peduncle, accessory flagellum absent; antenna

2 half as long as 1, flagellum 9-articulate; mandible (fig. 18B) with palp repre-

sented by a single seta, cutting edge strongly toothed, spine row of 10 spines,

the first three pectinate, molar absent; palp of maxilla 1 (Fig. 18C) uni-articulate,

tipped by seven spines, outer plate terminating in four strong serrate spines and

two small simple ones, inner plate bearing single apical seta; outer plate of

maxilla 2 (Fig. 18D) with four apical spines, inner plate bearing three setae;

inner plates of maxilliped (Fig. 18E) small, each bearing a single spine, outer

plate vestigial, palp 4-articulate.

Coxa 1 small, covered by coxa 2, which is produced anteriorly; coxa 3

rounded, bearing numerous chitinous (? stridulation) ridges along distal and

posterior margins; coxa 4 very large, not posteriorly excavate; remaining coxae

oval; gnathopod | small, subchelate, article 4 distally produced into a setose

lobe, article 5 as long as 6, palm oblique, undefined; gnathopod 2 extremely

large (Fig. 18A), antero-distal corners of articles 2 and 3 moderately lobed,

article 5 cup-shaped, 6 tapering distally from defining angle; palm greatly exceed-

THE AMPHIPODA OF SOUTHERN AFRICA 325

Fig. 18. Parametopa grandimana 0. sp.

Male, 3 mm: A —lateral aspect; B—mandible; C—maxilla 1; D—maxilla 2;

E—maxilliped; F—uropod 3; G—telson.

326 ANNALS OF THE SOUTH AFRICAN MUSEUM

ing hind margin, bearing about 12 small irregularly spaced teeth and scattered

small setae, defined by a larger tooth, dactyl powerful, equal to palm, inner

margin sparsely setose; pereiopods 1-3 slender, article 2 linear, article 4 widening

slightly distally; article 2 of pereiopods 4 and 5 oval, article 4 expanded distally

and postero-distally produced into an acute lobe bearing strong spines on both

margins.

Pleonal epimera | and 2 rounded, 3 rounded-quadrate; peduncle of uropod

1 with two rows of 7-8 dorsal spines, rami equal, lanceolate, bearing one or

two dorsal spines; uropod 2 half length of 1, outer ramus 70% length of inner,

each with a single dorsal spine; peduncle of uropod 3 (Fig. 18F) with three

dorsal spines, the single ramus consisting of two subequal articles, the first

bearing two apical spines, the second dorsally pectinate; telson (Fig. 18G)

entire, longer than broad, each lateral margin bearing one minute spine and

three powerful ones.

Holotype: SAM A13216, male, 3 mm.

Type locality: SST 5F, 35°22’'S/22°31’E, 20 June 1972, depth 200 m, substrate

coarse khaki sand.

Relationships: Of the three existing species in this genus P. grandimana n. sp.

can be distinguished from P. alaskensis (Holmes) and P. kervillei Chevreux by

the shape and size of gnathopod 2. I have been unable to obtain a description

of the third species—P. sarniensis (Norman)—but Chevreux & Fage, writing

after Norman, state that all members of Parametopa known at that time have

subequal antennae, a feature not consistent with the present species.

Material: SST 5F(10), SST 16K(6).

Proboloides rotunda (Stebbing, 1917)

Metopa rotundus Stebbing, 1917: 39, pl. 7A.

Records: SCD 122K(1), SCD 135L(2), SCD 159V(18), SCD 181Z(1), SCD

188F(2), SCD 198U(2), SCD 199E(2), SCD 204Y(1), SCD 211Z(1), SCD

222P(2), SCD 225N(1), SCD 345Y(3); 34°S/23°E/42 m (Stebbing 1917).

Distribution: Endemic to South Africa.

Stenothoe dolichopous K. H. Barnard, 1916

Stenothoe dolichopous K. H. Barnard, 1916: 153, pl. 26, figs 15-17.

Records: 32°S/28°E/176 m (K. H. Barnard 1916).

Diagnosis: Gnathopod 1 very elongate, articles 3 and 4 subequal, 5 equal to 2,

6 linear, shorter than 5, palm oblique, defined by two spines; articles 2 and 3

of gnathopod 2 antero-distally produced into rounded lobes, article 6 twice as

long as broad, palm occupying whole posterior margin, a conical tooth at its

centre, a second nearer the hinge and a large crenulate triangular process just

THE AMPHIPODA OF SOUTHERN AFRICA 327

before the hinge; third pleonal epimeron with a minute postero-distal tooth;

ramus of uropod 3 shorter than peduncle, second joint half length of first; telson

oval, each lateral margin bearing three spines.

Distribution: The above record is the only one to date.

Stenothoe gallensis Walker, 1904

Stenothoe gallensis: J. L. Barnard, 1971: 120, figs 62, 63.

Records: KNY 166J, KNY 176E.

Distribution: Cosmopolitan.

Stenothoe valida Dana, 1853

Stenothoe valida: Sivaprakasam, 1967a: 373, fig. 2a—b. J. L. Barnard, 19705: 250, fig. 165.

Records: SST 24N(1).

Distribution: Cosmopolitan in tropical and temperate seas.

Family Synopiidae

Tiron australis Stebbing, 1908

Tiron australis Stebbing, 1908a: 79, pl. 38.

Records: SCD 110T(1), SCD 131Z(2), SCD 173U(1), SCD 370T(1), SCD

373M(5); SST 34P(1), SST 37A(1), SST 37Y(1); 33°S/28°E/86 m (Stebbing

1908a).

Distribution: Endemic to Indian Ocean coast of South Africa.

Superfamily Talitroidea

(Revised J. L. Barnard 19725)

Family Ceinidae

Afrochiltonia capensis (K. H. Barnard, 1916)

Chiltonia capensis K. H. Barnard, 1916: 244, pl. 27, figs 38-40.

Afrochiltonia capensis: K. H. Barnard, 1955: 93.

Records: HAM 11C(C); HAV 5C(1), HAV 7L(P), HAV 18L(C); KNY 112C:

STJ 24E(A), STJ 27F(C).

Distribution: Endemic, Natal to Saldanha Bay.

Family Phliantidae

Palinnotus natalensis K. H. Barnard, 1940

Palinnotus natalensis K. H. Barnard, 1940: 445, fig. 22.

Records: J 11E; Y 12.

Distribution: India, east coast of South Africa.

328 ANNALS OF THE SOUTH AFRICAN MUSEUM

Plioplateia triquetra K. H. Barnard, 1916

Plioplateia triquetra K. H. Barnard, 1916: 156, pl. 26, figs 18-24.

Records: SCD 302Y(2), SCD 310F(1); 33°S/27°E/91 m (K. H. Barnard 1916).

Diagnosis: Body as broad as deep, rostrum upturned, bearing a tooth on each

side; antenna | half length of pereon; pereon segments 1-7 and pleon segments

1 and 2 each surmounted by a dorsal carina, that of pereon segment 1 deeply

bifid; each pereonite also bearing a flat horizontal backwardly-directed lateral

projection; maxilla 1 with palp; maxilliped bearing 4-articulate palp; gnathopods

subchelate; uropod 3 represented by an oval lobe-like peduncle, rami absent.

Distribution: Endemic to south coast of South Africa.

Temnophlias capensis K. H. Barnard, 1916

Temnophlias capensis K. H. Barnard, 1916: 158, pl. 26, figs 25-35.

Records: S 54G; SS 55B; Still Bay (K. H. Barnard 1940).

Distribution: Endemic, Still Bay to South West Africa.

Family Talitridae

Subfamily Hyalinae

Allorchestes inquirendus K. H. Barnard, 1940

Allorchestes inquirendus K. H. Barnard, 1940: 477, fig. 34b-c.

Records: MB 37L(1); Port Elizabeth (K. H. Barnard 1940).

Distribution: Endemic, Port Elizabeth to South West Africa.

Hyale grandicornis Kroyer, 1845

Hyale grandicornis: K. H. Barnard, 1955: 93, fig. 46. Hurley, 1957: 904-909, figs 1-23.

Records: KNY 166A, KNY 171C; J 11G; Q 7G; L 71; X 8A; K 8J; Y 12E;

T 3G; KN 2H; SS 55F; Port Elizabeth, East London (K. H. Barnard 1916);

Still Bay, Plettenberg Bay (K. H. Barnard 1940).

Distribution: Indo-Pacific, South Atlantic.

Hyale macrodactyla Stebbing, 1899

Hyale macrodactyla: Ledoyer, 1972: 273, fig. 77.

Records: H 9B.

Distribution: India. Madagascar, southern Africa.

THE AMPHIPODA OF SOUTHERN AFRICA 329

Hyale maroubrae Stebbing, 1899

Hyale maroubrae: Hurley, 1957: 913, figs 51-71.

Records: KNY 166H(1), KNY 171D.

Diagnosis: Antenna | extending 30% along length of flagellum of antenna 2;

coxae rectangular; palm of gnathopod 1 male transverse, defining angle ridged;

article 2 of gnathopod 2 male not distally lobed, palm extending whole length

of article 6, bearing two rows each of eight seta-tipped spines and defined by a

shallow pocket into which the dactyl closes; article 6 of pereiopods 1—5 postero-

distally bearing a stout striated seta-tipped spine and a flattened fusiform striated

spine.

Distribution: Indo-Pacific.

Hyale saldanha Chilton, 1912

Hyale saldanha Chilton, 1912: 509, pl. 2, figs 24-29.

Records: MB 40N(2); L 38Z; E 232; AR 1Q(8).

Distribution: Endemic, East London to South West Africa.

Subfamily Talitrinae

Orchestia ancheidos (K. H. Barnard, 1916)

Talorchestia ancheidos K. H. Barnard, 1916: 221, pl. 27, figs 35, 36.

Orchestia ancheidos: Ruffo, 1958: 43, figs 3, 4.

Records: BMR 21E(6); HAM 4B(A); HAV 13M(1); STJ 5B(C), STJ 29F;

Keurbooms River, Plettenberg Bay (K. H. Barnard 1940).

Distribution: Madagascar, Mocgambique, South Africa.

Orchestia rectipalma K. H. Barnard, 1940

Parorchestia tenuis (non Dana, 1853): K. H. Barnard, 1916: 226.

Parorchestia rectipalma K. H. Barnard, 1940: 473, fig. 32.

Records: HAM 11B(C), HAM 13C(C); HAV 7J(P), HAV 18E(C), HAV 20A(C);

KNY 103A, KNY 175E, KNY 179A(C), KNY 181A(A), KNY 266C(5),

KNY 272A(18), KNY 273B(60), KNY 274H(4), KNY 285A(C), KNY 286B(C),

KNY 294A(47); BRE 13D(3), BRE 34K(4), BRE 55E(A), BRE 56F(1), BRE

57G(4), BRE 71N, BRE 123D(8); STJ 7Q, STJ 15N(C); Keurbooms River,

Plettenberg Bay (K. H. Barnard 1940).

Distribution: Endemic, Natal to South West Africa, especially in estuaries.

Talorchestia australis K. H. Barnard, 1916

Talorchestia australis K. H. Barnard, 1916: 220, pl. 27, figs 33, 34; 1940, fig. 30.

Records: GBR 13D(\).

Distribution: Endemic, South West Africa to Mocambique.

330 ANNALS OF THE SOUTH AFRICAN MUSEUM

Talorchestia capensis (Dana, 1853)

Talorchestia capensis; K. H. Barnard, 1916: 216; 1940: 470, fig. 28.

Records: KNY 43A, KNY 114A(C), KNY 166B, KNY 187A; BRE 5A(6),

BRE 18A(10); STJ SA(C); Keurbooms River, Plettenberg Bay (K. H. Barnard

1940).

Diagnosis: Male gnathopod 1 with article 5 not lobed distally, article 6 weakly

lobed; palm of gnathopod 2 male oblique, spinose, medially with a semicircular

incision which in adults occupies almost the entire palm, dactyl with a deep

semicircular concavity near its base; pleonal epimera postero-distally rounded,

their posterior margins distinctly crenulate.

Distribution: Mediterranean, Atlantic, south coast of South Africa.

Talorchestia inaequalipes K. H. Barnard, 1951

Talorchestia inaequalipes K. H. Barnard, 1951: 705, fig. Sa—b.

Records: KNY 113A(C), KNY 162B.

Diagnosis: Eyes dorsally separated by more than their diameter; coxa 2 not

lobed posteriorly; articles 5 and 6 of gnathopod | male apically lobed; palm of

gnathopod 2 male oblique, spinose, defined by a small pelucid lobe; article 2

of pereiopod 5 weakly serrate posteriorly, articles 4 and 5 strongly expanded in

male, oar-like; hind margins of pleonal epimera distinctly serrate, postero-

distal corners quadrate.

Distribution: Endemic, Knysna to Saldanha Bay.

Suborder CAPRELLIDEA

Family Aeginellidae

Eupariambus fallax K. H. Barnard, 1957

Eupariambus fallax K. H. Barnard, 1957: 9, fig. 6.

Records: SST 29D(2), SST 45E(3).

Diagnosis: Antenna 1 without swimming setae; mandible with molar and

3-articulate palp; branchiae on segments 3 and 4; palm of gnathopod 2 with a

large acute tooth at its centre; pereiopods | and 2 absent, pereiopod 3 reduced

to a 2-articulate rudiment terminating in four or five setae; abdomen of male

with one pair of rudimentary appendages.

Distribution: Endemic, Still Bay to west coast of South Africa.

Metaprotella macrodactylos Stebbing, 1910

Metaprotella macrodactylos Stebbing, 1910a: 469, pl. 48A.

Records: 33°S/26°E/18-29 m (Stebbing 1910a).

Distribution: Endemic, Natal to Port Elizabeth.

THE AMPHIPODA OF SOUTHERN AFRICA 331

Orthoprotella mayeri K. H. Barnard, 1916

Orthoprotella mayeri K. H. Barnard, 1916: 284; 1925: 372.

Records: SCD 3U(2), SCD 60M(3), SCD 122J(2), SCD 160L(2), SCD 181G(11),

SCD 204N(4), SCD 319Y(3); SST 11T(7), SST 18W(1), SST 24B(4), SST

37Q(2); 33°S/27°E/100 m, 34°S/22°E/86 m (K. H. Barnard 1916); Algoa Bay

184 m, 33°S/28°E/180 m, 34°S/25°E/133 m, 34°S/23°E/148 m (K. H. Barnard

1925).

Distribution: Indo-Pacific.

Pseudaeginella tristanensis (Stebbing, 1888)

Pseudaeginella tristanensis: Stephensen, 1949: 52, fig. 23.

Records: SCD 160K(1), SCD 181Z; L 455; East London (K. H. Barnard 1940).

Distribution: Tristan da Cunha, South Africa.

Family Caprellidae

Caprella cicur Mayer, 1903

Caprella cicur Mayer, 1903: 75, 97, pl. 4, figs 5-7, pl. 8, figs 3-5.

Records: SCD 202H(1), SCD 248L(1); J 11C; LLL 6C; Port Elizabeth 24-27 m,

33°S/26°E/18-29 m (Stebbing 1910a).

Distribution: Endemic, Natal to west coast of South Africa.

Caprella danilevskii Czerniavski, 1868

Caprella danilevskii: McCain, 1968: 22-25, figs 10, 11.

Records: J 11C; L 500; K 8H; E 236.

Distribution: Widespread, pantropical.

Caprella equilibra Say, 1818

Caprella equilibra: McCain, 1968: 25-30, figs 12, 13.

Records: SCD 179K(1), SCD 189T(6), SCD 192W(65), SCD 198G(1), SCD

244E(9), SCD 282U(4), SCD 283F(1), SCD 312N(35), SCD 329Y(1), SCD

338J(5), SCD 353L(1), SCD 379H(6); MB 37K(6), MB 54W(1); KNY 28B;

Port Elizabeth (K. H. Barnard 1916).

Distribution: Cosmopolitan.

Caprella natalensis Mayer, 1903

Caprella penantis (non Leach, 1814): Stebbing, 1910a: 465.

Caprella natalensis: Laubitz, 1972: 47, pl. 9, figs F, G, pl. 10, figs F—K.

Records: Port Elizabeth 24-27 m (Stebbing 1910a).

332 ANNALS OF THE SOUTH AFRICAN MUSEUM

Distribution: Southern Africa, Tristan da Cunha, Pacific coast of North America.

Caprella penantis Leach, 1814

Caprella penantis: McCain, 1968: 33-40, figs 15, 16.

Records: MB 413(1): D272; tb 335, 455B; Y 12D.

Distribution: Cosmopolitan in tropical and temperate seas.

Caprella scaura Templeton, 1836

Caprella scaura: McCain, 1968: 40-44, figs 17, 18.

Records: MB 50U(S).

Distribution: Cosmopolitan.

Caprella triodous Stebbing, 1910

Caprella triodous Stebbing, 1910a: 467, pl. 48B.

Records: SCD 141W(1); 33°S/25°E/24-27 m (Stebbing 1910a).

Diagnosis: Head without rostrum; flagellum of antenna 1 12-articulate; antenna

2 shorter than peduncle of antenna 1; no spine between bases of second gnatho-

pods; article 6 of gnathopod 2 elongate, distally widening, palm divided into

three very large teeth; pereiopods 3 and 4 each with a pair of serrate-ended

clasping spines.

Distribution: Endemic to south coast of South Africa.

Hemiaegina minuta Mayer, 1890

Hemiaegina minuta: McCain, 1968: 61-64, figs 29, 30.

Records: SCD 179L(1); L 61A.

Distribution: Cosmopolitan in tropical and temperate seas.

Family Phtisicidae

Caprellina longicollis (Nicolet, 1849)

Caprellina longicollis: McCain, 1969; 289, fig. 2.

Records: L 455A; 33°S/26°E/18-29 m, Port Elizabeth 24-27 m (Stebbing 1910a).

Distribution: Mediterranean, southern oceans.

Caprellina spiniger K. H. Barnard, 1916

Caprellina spiniger K. H. Barnard, 1916: 282, pl. 28, fig. 35.

Records: MB S0V(1).

Distribution: Endemic, Mossel Bay to Liideritz.

THE AMPHIPODA OF SOUTHERN AFRICA 333

Chaka leoni Griffiths, 1974

Chaka leoni Griffiths 19746: 258, figs 7, 8.

Records: SCD 59E(1).

Distribution: Endemic to east and south coasts of South Africa.

Phtisica marina Slabber, 1769

Phtisica marina: K. H. Barnard, 1916: 283. McCain, 1968: 91-97, fig. 46.

Records: SCD 24P(1), SCD 135H(1), SCD 159T(4), SCD 160J(2), SCD 181H(7),

SCD 188E(3), SCD 192X(10), SCD 216J(1), SCD 232G(4), SCD 345Z(2),

SCD 347C(1); SST 11S(13), SST 29C(1); 34°S/23°E/84 m (K. H. Barnard 1916).

Distribution: Mediterranean, Black Sea, Atlantic, southern Africa.

SUMMARY

An account is presented of the known gammaridean and caprellid amphipod

fauna of the Cape Province of South Africa east of 20°E (Cape Agulhas). The

vast majority of the records has been derived from the extensive estuarine,

littoral and benthic collections amassed by the University of Cape Town

Ecological Survey, and with these have been incorporated all previous records

from the literature. The collections total in excess of 12 000 individuals and from

these 173 species have been recognized. Fourteen of the species and two genera

are presented here as new to science. These are Dikwa acrania n. gen., n. sp.

(Acanthonotozomatidae); Ampelisca acris n. sp.; Colomastix keiskama Nn. sp.;

Concholestes armatus n. sp.; Neomicrodeutopus nyala n. sp.; Cunicus profundus

n. gen., n. sp. (Haustoriidae); Urothoe platypoda n. sp.; Parajassa chikoa n. sp.;

Liljeborgia palmata n. sp.; Listriella sinuosa n. sp.; Lepidepecreum twalae n. sp.;

Uristes sulcus n. sp.; Westwoodilla manta n. sp. and Parametopa grandimana

n. sp. In addition two species, Stomacontion prionoplax Monod and Podocerus

hystrix Stebbing, are recorded for the first time from southern Africa. A fusion

of the Lysianassid genera Stomacontion and Acontiostoma is advocated.

ACKNOWLEDGEMENTS

My thanks go to Professor J. H. Day under whose constructive guidance

this work was carried out; also to Mr B. F. Kensley of the South African Museum

for the loan of type material, and to Miss Belle Leon for her help in the mono-

tonous tasks of cataloguing and proof reading. This project would not have

been possible but for the financial assistance generously provided by the South

African Council for Scientific and Industrial Research.

334 ANNALS OF THE SOUTH AFRICAN MUSEUM

REFERENCES

BARNARD, J. L. 1954. Marine Amphipoda of Oregon. Ore. St. Monogr. Stud. Zool. 8: 1-103.

BARNARD, J. L. 1955. Gammaridean Amphipoda (Crustacea) in the collections of the Bishop

Museum. Bull. Bernice P. Bishop Mus. 215: 1-46.

BARNARD, J. L. 1957. A new genus of phoxocephalid Amphipoda (Crustacea) from Africa,

India, and California. Ann. Mag. nat. Hist. (12) 10: 432-438.

BARNARD, J. L. 1959a. Liljeborgiid amphipods of southern California coastal bottoms, with

a revision of the family. Pacif. Nat. 1: 12-28.

BARNARD, J. L. 1959b. The common pardaliscid Amphipoda of southern California, with a

revision of the family. Pacif. Nat. 1: 36-43.

BARNARD, J. L. 1960. The amphipod family Phoxocephalidae in the eastern Pacific Ocean,

with analyses of other species and notes for a revision of the family. Allan Hancock

Pacif. Exped. 18: 175-368.

BARNARD, J. L. 1961. Gammaridean Amphipoda from depths of 400 to 6 000 meters. Galathea

Rep. 5: 23-128.

BARNARD, J. L. 1962. Benthic marine Amphipoda of southern California: Families Aoridae,

Photidae, Ischyroceridae, Corophiidae, Podoceridae. Pacif. Nat. 3: 1-72.

BARNARD, J. L. 1964. Los anfipodos bentonicos marinos de la costa occidental de Baja

California. Revta Soc. mex. Hist. nat. 24: 205-274.

BARNARD, J. L. 1965. Marine Amphipoda of atolls in Micronesia. Proc. U.S. natn. Mus.

117: 459-552.

BARNARD, J. L. 1969a. Gammaridean Amphipoda of the rocky intertidal of California:

Monterey Bay to La Jolla. Bull. U.S. natn. Mus. 258: 1-230.

BARNARD, J. L. 19696. The families and genera of marine gammaridean Amphipoda. Bull.

U.S. natn. Mus. 271: 1-535.

BARNARD, J. L. 1970a. The identity of Dexamonica and Prinassus with a revision of Dexamini-

dae (Amphipoda). Crustaceana 19: 161-180.

BARNARD, J. L. 19706. Sublittoral Gammaridea (Amphipoda) of the Hawaiian Islands.

Smithson. Contr. Zool. 34: 1-286.

BARNARD, J. L. 1971. Keys to the Hawaiian marine Gammaridea, 0-30 meters. Smithson.

Contr. Zool. 58: 1-135.

BARNARD, J. L. 1972a. Gammaridean Amphipoda of Australia, part 1. Smithson. Contr.

Zool. 103: 1-333.

BARNARD, J. L. 19726. The marine fauna of New Zealand: Algae-living littoral Gammaridea.

(Crustacea Amphipoda). Mem. N.Z. oceanogr. Inst. 62: 1-216.

BARNARD, K. H. 1916. Contributions to the crustacean fauna of South Africa. 5. The Amphi-

poda. Ann. S. Afr. Mus. 15: 105-302.

BARNARD, K. H. 1925. Contributions to the crustacean fauna of South Africa. 8. Further

additions to the list of Amphipoda. Ann. S. Afr. Mus. 20: 319-380.

BARNARD, K. H. 1932. Amphipoda. Discovery Rep. 5: 1-326.

BARNARD, K. H. 1935. Report on some Amphipoda, Isopoda and Tanaidacea in the collection

of the Indian Museum. Rec. Indian Mus. 37: 279-319.

BARNARD, K. H. 1937. Amphipoda. Scient. Rep. John Murray Exped. 1933-34. 4: 131-201.

BARNARD, K. H. 1940. Contributions to the crustacean fauna of South Africa. XII. Further

additions to the Tanaidacea, Isopoda, and Amphipoda with keys for the identification

of hitherto recorded marine and fresh-water species. Ann. S. Afr. Mus. 32: 381-543.

BARNARD, K. H. 1951. New records and descriptions of new species of isopods and amphipods

from South Africa. Ann. Mag. nat. Hist. (12) 4: 698-709.

BARNARD, K. H. 1955. Additions to the fauna-list of South African Crustacea and Pycnogonida.

Ann. S. Afr. Mus. 43: 1-107.

BARNARD, K. H. 1957. Additions to the fauna-list of South African Crustacea. Ann. Mag. nat.

Hist. (12) 10: 1-12.

BousFIELD, E. L. 1970. Adaptive radiation in sand-burrowing amphipod crustaceans. Chesa-

peake Sci. 11: 143-154.

CHEVREUX, E. & Fage, L. 1925. Amphipodes. Faune Fr. 9: 1-488.

CHILTON, C. 1912. The Amphipoda of the Scottish National Antarctic Expedition. Trans. R.

Soc. Edinb. 48: 455-520.

CHILTON, C. 1921. Fauna of the Chilka Lake, Amphipoda. Mem. Indian Mus. 5: 521-556.

THE AMPHIPODA OF SOUTHERN AFRICA 335

Dana, J. D. 1853. Crustacea, Part 2. U.S. Expl. Exped. 14: 689-1618.

Day, J. H., MILLARD, N. A. H. & HARRISON, A. D. 1952. The ecology of South African

estuaries. Part 3. Knysna: A clear open estuary. Trans. R. Soc. S. Afr. 34: 367-413.

GILES, G. M. 1888. Natural history notes from H.M.’s Indian Marine Survey Steamer ‘Investi-

gator’, Commander Alfred Carpenter, R.N., D.S.O., commanding. No. 9: Further notes

on the Amphipoda of Indian waters. J. Asiat. Soc. Beng. 59: 63-74.

Gray, W. S. & BARNARD, J. L. 1970. South African Ampelisca excavata K. H. Barnard

(Amphipoda, Gammaridea), a redescription with notes on the domicile. Crustaceana

19: 67-83.

GRIFFITHS, C. L. 1973. The Amphipoda of Southern Africa, Part 1. The Gammaridea and

Caprellidea of Southern Mocambique. Ann. S. Afr. Mus. 60: 265-306.

GriFFiTus, C. L. 1974a. The Amphipoda of Southern Africa, Part 2. The Gammaridea and

Caprellidea of South West Africa south of 20°S. Ann. S. Afr. Mus. 62: 169-208.

GRIFFITHS, C. L. 19746. The Amphipoda of Southern Africa, Part 3. The Gammaridea and

Caprellidea of Natal. Ann. S. Afr. Mus. 62: 209-264.

HUuRLEY, D. E. 1957. Studies on the New Zealand amphipodan fauna. No. 14. The genera

Hyale and Allorchestes (Family Talitridae). Trans. R. Soc. N.Z. 84: 903-933.

KAIM MALKA, R. A. 1969. Contribution a l’étude de quelques espéces du genre Ampelisca

(Crustacea-Amphipoda) en Méditerranée. II. Téthys 1: 927-977.

LausiTZz, D. R. 1972. The Caprellidae (Crustacea, Amphipoda) of Atlantic and Arctic Canada.

Publs biol. Oceanogr. natn. Mus. nat. Sci. (Can.) 4: 1-82.

LEDOYER, M. 1967. Amphipodes gammariens des herbiers de phanérogames marines de la

région de Tuléar (République Malgache). Etude systématique et écologique. Annls

Univ. Madagascar (Sci. nat. Math.) 5: 121-170.

LEDOYER, M. 1972. Amphipodes gammariens vivant dans les alveoles des constructions

organogenes récifales intertidales de la région de Tuléar (Madagascar). Etude systématique

et écologique. Téthys, Suppl. 3: 165-286.

MAYER, P. 1903. Die Caprellidae der Siboga Expedition. Siboga Exped. 34: 1-158.

McCain, J. C. 1968. The Caprellidae (Crustacea: Amphipoda) of the western North Atlantic.

Bull. U.S. natn. Mus. 278: 1-147.

McCain, J. C. 1969. New Zealand Caprellidae (Crustacea: Amphipoda). N.Z. J/ mar. freshw.

Res. 3: 286-295.

MILs, E. L. 1971. Deep-sea Amphipoda from the western North Atlantic Ocean. The family

Ampeliscidae. Limnol. Oceanogr. 16: 357-387.

Monon, Tu. 1937. Missions A. Gruvel dans le canal de Suez. I. Crustacés. Mém. Inst. Egypte

34: 1-19.

Myers, A. A. 1970. Taxonomic studies on the genus Grandidierella Coutiére (Crustacea:

Amphipoda), with a description of G. dentimera sp. nov. Bull. mar. Sci. Gulf Caribb.

20: 135-147.

NaGartTa, K. 1965. Studies on marine gammaridean Amphipoda of the Seto Inland Sea. III.

Publs Seto mar. biol. Lab. 13: 291-326.

PILLAI, N. K. 1957. Pelagic Crustacea of Travancore, III: Amphipoda. Bull. cent. Res. Inst.

Univ. Travancore (C) 5: 29-68.

RABINDRANATH, P. 1971. Haustoriid amphipods (Crustacea) from India. Hydrobiologia 38:

521-539.

Rew, D. M. 1951. Report on the Amphipoda (Gammaridea and Caprellidea) of the coast of

tropical West Africa. Atlantide Rep. 2: 189-291. ;

RuFFO, S. 1958. Amphipodes terrestres et des eaux continentales de Madagascar, des Comores

et de la Réunion. Mém. Inst. scient. Madagascar 12: 35-64.

RuFFo, S. 1969. Studi sui crostacei amfipodi LXVII. Terzo contributo alla conscenza delgi

amfipodi del mar Rosso. Memorie Mus. civ. Stor. nat. Verona 17: 1-77.

Sars, G. O. 1895. An account of the Crustacea of Norway, with short descriptions and figures

of all the species. 1. Amphipoda. Christiania, Copenhagen: Cammermeyers.

SCHELLENBERG, A. 1925. Crustacea VIII: Amphipoda. Beitr. Kennt. Meeresfauna Westafr.

3: 113-204.

SCHELLENBERG, A. 1953. Erganzungen zur Amphipoden-fauna Siidwest-Africas nebst Bemer-

kungen tiber Brutraumbildung. Mitt. zool. Mus. Berl. 29: 107-126.

SEXTON, E. W. & Rep, D. M. 1951. The life-history of the multiform species Jassa falcata

336 ANNALS OF THE SOUTH AFRICAN MUSEUM

(Montagu) (Crustacea, Amphipoda) with a review of the bibliography of the species.

J. Linn. Soc. (Zool.) 42: 29-91.

SIVAPRAKASAM, T. E. 1966. Amphipoda from the east coast of India. Part 1. Gammaridea.

J. mar. biol. Ass. India 8: 82-122.

SIVAPRAKASAM, T. E. 1967a. Notes on some amphipods from the south east coast of India.

J. mar. biol, Ass. India 9: 372-383.

SIVAPRAKASAM, T. E. 1967b. Leucothoid Amphipoda from the Madras coast. J. mar. biol.

Ass. India 9: 384-391.

SIVAPRAKASAM, T. E. 1968. A new species and a new record of Amphipoda from the Madras

coast. J. mar. biol. Ass. India 10: 274-282.

SIVAPRAKASAM, T. E. 1969. Amphipoda from the east coast of India. 2. Gammaridea and

Caprellidea. J. Bombay nat. Hist. Soc. 66: 560-576.

STEBBING, T. R. R. 1888. Report on the Amphipoda collected by H.M.S. Challenger during

the years 1873-76. Rep. Voy. Challenger 1873-76 (Zool.) 29: i-xxiv, 1-1737.

STEBBING, T. R. R. 1904. Gregarious Crustacea from Ceylon. Spolia zeylan. 2: 1-29.

STEBBING, T. R. R. 1906. Amphipoda 1. Gammaridea. Tierreich 21: i-xxxix, 1-806.

STEBBING, T. R. R. 1908a. South African Crustacea (Part IV). Ann. S. Afr. Mus. 6: 1-96.

STEBBING, T. R. R. 19085. The fauna of brackish ponds at Port Canning, Lower Bengal.

Part 9: A new species of Amphipoda. Rec. Indian Mus. 2: 119-123.

STEBBING, T. R. R. 1910a. General catalogue of South African Crustacea (Part V of S.A.

Crustacea). Ann. S. Afr. Mus. 6: 447-473.

STEBBING, T. R. R. 19105. Scientific results of the trawling expedition of H.M.C.S. ‘Thesis’

off the coast of New South Wales in February and March, 1898. Crustacea. Part 5.

Amphipoda. Mem. Aust. Mus. 4: 565-658.

STEBBING, T. R. R. 1917. South African Crustacea (Part IX). Ann. S. Afr. Mus. 17: 23-46.

STEBBING, T. R. R. 1918. Some Crustacea of Natal. Ann. Durban Mus. 2: 47-75.

STEPHENSEN, K. 1949. The Amphipoda of Tristan da Cunha. Results Norw. scient. Exped.

Tristan da Cunha 3 (19): 1-61.

VADER, W. 1970. The status of Bathyporeia gracilis Sars (Amphipoda, Haustoriidae). Sarsia

43: 155-162.

WALKER, A. O. 1904. Report on the Amphipoda collected by Professor Herdman, at Ceylon,

in 1902. In: ROYAL SOCIETY. Report to the government of Ceylon on the pearl oyster fisheries

of the gulf of Manaar, by W. A. Herdman. Part II (suppl. rep. 17): 229-300. London: the

Society.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

[ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FISCHER, P.-H., DUvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Koun, A. J. 19605. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270. 3

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

C. L. Griffiths

THE AMPHIPODA OF SOUTHERN AFRICA

PART 4

THE GAMMARIDEA AND CAPRELLIDEA OF

THE CAPE PROVINCE EAST OF CAPE AGULHAS

5°7,68

VOLUME 65 PART 10 OCTOBER 1974

ANNALS

| CAPE TOWN

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 ~~ Band

October 1974 Oktober

Part 10 Deel

THE CRANIAL MORPHOLOGY OF

THRINAXODON LIORHINUS SEELEY

S. FOURIE

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1,24, 3; 5, 7=8); 3C1-2, 5, t-=p.1.), 5(1—2, 5; 7-9);

6(1, t.-p.i.), 711-3), 8, 911-2), 10(1),

11(1-2, 5, 7, t.—p.i.), 24(2), 27, 31(1-3), 33

Price of this part/Prys van hierdie deel

R8,40

1974

ISBN 0 949940 50 X

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

THE CRANIAL MORPHOLOGY OF

THRINAXODON LIORHINUS SEELEY

S. FOURIE

Department of Zoology, University of the O.F.S., Bloemfontein

(With 33 figures)

LMS. accepted 22 August 1973]

CONTENTS

PAGE

Introduction : : : ; : : aes

Description of material

(a) Skull. : ; F : : : oh a Sat

(b) Lower jaw . : ; : . in S84

Summary . : ; . ‘ E : BBS)

Acknowledgements . : 5 : 5 eS 94:

References . : ‘ at Pi : : . 394

Key to lettering . : : ; : 3 oo

INTRODUCTION

When Owen introduced the term “Cynodontia’ in 1861, he made Galesaurus,

which has a mammal-like dentition, the type of the new group which he called

a ‘family’ of the order Anomodontia. Only in 1876 did he create the order

Theriodontia to receive all forms with a carnivorous mammal-like dentition.

Lycosaurus, Nythosaurus and Scaloposaurus were included in this order. The

former, which was the type of the order, has simple, pointed teeth whereas the

two other genera have laterally cuspidate postcanines. Obviously the Therio-

dontia were therefore meant by Owen to include more than his Cynodontia.

Because the crowns of the postcanines were not preserved in the Galesaurus

specimen originally used by Owen (1861) as the type of his Cynodontia, Seeley

(18955: 59) proposed that the name Cynodontia be used for animals which

resemble Galesaurus in skull structure and Nythosaurus in the structure of the

postcanines, because it was possible to distinguish the Cynodontia from the

Lycosauria (Theriodontia) ‘. . . by dental and other minor characters of the

skull’. Seeley therefore designated Cynognathus, which he described in that year,

as the type of the more narrowly defined Cynodontia. In 1903 Broom pointed

out that the Theriodontia was not a natural order and proposed that it be

broken down into Cynodontia and Therocephalia. Among these Cynodontia

none is better known today than Thrinaxodon.

The first skull of Thrinaxodon was described by Owen in 1887, but he

referred it to Galesaurus planiceps. In this identification he was followed by

Seeley (1889) and Lydekker (1890). Only in 1895, on comparing the postcanine

crown patterns of Galesaurus planiceps and Nythosaurus larvatus did Seeley

337

Ann. S. Afr. Mus. 65 (10), 1974: 337-400, 33 figs.

338 ANNALS OF THE SOUTH AFRICAN MUSEUM

(1895a) realize that the skull belonged to a new genus and species for which he

proposed the name Thrinaxodon liorhinus. Van Hoepen (1916), Gregory (1920)

and Haughton (1920, 1924), when describing specimens now recognized as

members of the genus Thrinaxodon, referred them to Ictidopsis, a name intro-

duced by Broom in error in 1912. Since these two names refer to identical forms,

the term Thrinaxodon, introduced by Seeley in 1894, has preference.

In 1932 Broom added the species T. putterilli to Seeley’s original T. liorhinus.

Thrinaxodon has become the best known of the Galesauridae and Cynodontia

as a result of the studies of Broom published in several papers between 1910

and 1938, Watson (1920), Parrington (1933, 1935, 1936, 1939, 1946), Olson

(1944), and Brink (1954, 19555). More recently Estes (1961) elucidated the cranial

structure of Thrinaxodon with emphasis on young individuals. Crompton

(1963a) and Osborn & Crompton (1973) explained tooth replacement in

Thrinaxodon and its significance in relation to mammalian evolution. Crompton

& Jenkins (1968) demonstrated the importance of the dentition of Thrinaxodon

in the evolution of molar teeth and the development of molar occlusion in the

first mammals of the late Triassic. Hopson (1966), in a discussion of the origin of

the mammalian middle ear, described the condition in Thrinaxodon and noted

that the functional aspects of the cynodont middle ear are not understood well

enough to say much about acuteness of hearing, but the largeness of the fenestra

ovalis relative to the apparent size of the tympanum, which had an area of

one-ninth or less of that of the former, would probably have made the sensitivity

to sound much lower than in living reptiles.

Van Heerden (1972) has just completed a revision of some of the Galesauri-

dae. He demonstrated the identity of Thrinaxodon putterilli with Thrinaxodon

liorhinus. The type specimens of Notictosaurus luckhoffii and N. gracilis, declared

identical with each other by Brink (1965), have now been shown to be identical

with T. liorhinus. Van Heerden’s investigations have greatly extended the terri-

torial distribution of Thrinaxodon between Harrismith and vicinity in the north-

eastern Orange Free State and Thaba N’chu more west and south, much further

south to the vicinity of the Hendrik Verwoerd Dam near Bethulie, then on to

the vicinity of Burgersdorp slightly more south-east and to Graaff-Reinet in

the far south-west. Finally the value of the detailed description by Jenkins

(1971) of the postcranial anatomy of Thrinaxodon and other African cynodonts

should be mentioned in this brief survey of important contributions to our

knowledge of Thrinaxodon.

In 1938 Broom had a skull sectioned at intervals of 4 to 5 mm. Considerable

detail of internal structure and some of tooth replacement was obtained. The

intervals between sections were, however, large and much detail was lost.

Consequently accurate interpretations were difficult and several minor errors

slipped in. Broom himself realized the inadequacy and suggested that a much

larger series of sections and photographs be made in order ‘. . . to clear up all

the characters of the prootic and opisthotic and to reveal the structure of the

labyrinth’. Some of this additional information was obtained by Olson (1944)

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 339

when he sectioned a skull at intervals of 0,37 mm. He concentrated on the

braincase, otic region and base of the skull.

Parrington’s paper of 1946 added much new information, but it was felt

that serial grinding at closer intervals than those taken in previous investigations

might settle doubtful points, and reveal additional information especially on

tooth replacement, the course of some of the nerves and blood vessels, the

middle and inner ear, the relationships of the elements forming the braincase,

the relationships of the postdentary bones, and the origins and insertions of

jaw muscles and correlated skull structure. With this purpose in mind the study

on which the present paper is based was started.

The skull of Thrinaxodon liorhinus Seeley selected for sectioning measured

67 mm from the tip of the snout to the posterior termination of the parietal

crest. It was catalogued as number C354 of the National Museum, Bloemfon-

tein, and was very well preserved, showing hardly any distortion. The snout was

slightly damaged and the right postorbital arch was incomplete. Several detached

pieces of bone lodged in the matrix were possibly derived from the broken

regions. Serial grinding was done at intervals of 0,2 mm. Enlarged drawings

were used for making graphic reconstructions and for building a wax plate

model (5 x, linear) of the skull. Crompton’s (1955a) modified method based

on the original technique described by Sollas (1914) and Olson (1944) was used.

Photographs were taken of sections featuring important detail.

Before the completion of the original work on which this paper is based,

the paper by Estes (1961) appeared which shed further light, especially on

growth changes in the Thrinaxodon skull.

Through these studies the taxonomic position and external features of the

skull and skeleton of Thrinaxodon have become well known. During the last

five years it has become more and more evident that Thrinaxodon was an animal

of crucial importance in the evolution of the first mammals.

In his discussion of the lower jaw of cynodonts and the evolutionary origin

of the mammal-like adductor jaw musculature Barghusen (1968) showed that

in Thrinaxodon a muscle mass with relationships approaching those of the

mammalian masseter muscle had developed. The development of this muscula-

ture is intimately related to and can account for the relative posterior expansion

of the ventral portion of the dentary ramus. This extra surface was included by

the masseter muscle. Contrary to Crompton’s (1963b) view, Barghusen shows

that the adductor musculature did not migrate from the reflected lamina on to

the ventral part of the masseteric fossa, but that the external adductor muscle

(Crompton’s deep masseter) on the dorsal lateral part of the dentary in pro-

cynosuchids migrated downwards (ventrally) and at the same time the entire

length of the zygomatic arch was established as an area of muscular origin.

The cynodont masseter is therefore a new differentiation not found in other

therapsids.

The second change in cynodonts was the development of a pattern of inser-

tion characteristic of the mammalian temporalis muscle. This was accomplished

340 ANNALS OF THE SOUTH AFRICAN MUSEUM

through the posterodorsal expansion of the coronoid process of the dentary.

As a result of these modifications cynodonts established an arrangement of

adductor jaw musculature closely approaching that in living mammals. Cromp-

ton (19635) had shown that in a stratigraphic series of cynodont lower jaws

progressive osteological changes not found in any other group of reptiles took

place. These changes were the progressive enlargement of the posterior part of

of the dentary and concomitantly the reorientation in position and reduction in

size of the accessory jawbones. Both Crompton and Barghusen relate the osteo-

logical changes in the lower jaw to the development of the mammalian type jaw

musculature. Thus differentiation of jaw musculature in early cynodont evolu-

tion (procynosuchids to Thrinaxodon) is the key innovation that had profound

significance for the origin of mammals.

In 1967 Hopson, in a discussion of mammal-like reptiles and the origin of

mammals, concluded that early mammals could all be derived from small,

probably insect-eating early Triassic cynodonts with skull and dental characters

like those of Thrinaxodon. Some indication of the possible continued existence

in the late Triassic of persistently small, dentally conservative cynodonts which

were acquiring mammalian features in parallel with their larger more specialized

relatives is a single poorly preserved skull of a very small cynodont with

Thrinaxodon-like teeth from the Middle Triassic.

On the basis of the fact that the crown structure of one of the generations

of postcanine teeth of Thrinaxodon is almost identical to that of Eozostrodon,

except that the Thrinaxodon postcanines have a single root, and due to the

overall primitive structure of the primitive cynodont skull, Crompton & Jenkins

(1968) concluded that Rhaetic mammals (excluding ictidosaurs and haramyids)

could be derived from primitive cynodonts. They point out that new material

shows that it is possible to derive the cynodont petrosal and alisphenoid from

that of a scaloposaurid, and that the eozostrodontid petrosal, as it is known at

present, can be derived from that of an early cynodont of which the anterior

margin is more complex than is shown in published accounts.

The view that mammals can be derived from persisting members of the

cynodont family Galesauridae, of which Thrinaxodon is the most important

known member, is reiterated by Hopson & Crompton (1969). They state that

only cynodonts possess a large suite of features elsewhere found only in mam-

mals. They believe that mammals had a monophyletic origin at the low taxono-

mic level of a family or even lower.

At present active research is being conducted on late cynodonts of East

Africa and South America, in which information on the transitional stages

between the early South African cynodonts and the first mammals is sought.

In spite of all past and recent work, problems of functional anatomy and

evolutionary transformation still remain. It is hoped that the results described

in the present paper might, though obtained about ten years ago as part of the

work for a doctoral thesis, still be able to contribute to the detailed comparison

of structure in early and later cynodont and early mammalian skulls.

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 34]

Fig. 1. Graphic reconstruction of ventral view of serially sectioned

skull shown as a mirror image. x 1,5.

DESCRIPTION OF MATERIAL

(a) SKULL (Figs 1, 2)

In this section only new details that have appeared from the present study

will be mentioned.

Premaxillary (Figs 1, 3-6)

Behind the first incisor in each premaxillary is the small foramen (Figs

1, 33C, FNNG) of a vertical canal which may have had some connection with

an organ of Jacobson or a nasal gland connected with the ability to

detect odours. The canal appears to have a horizontal branch which opens

on the anterior surface of the premaxillary immediately below the base of

the internarial process. The vertical canal apparently opens on the floor

342 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 2. Dorsal view of model of serially sectioned skull of Thrinaxodon liorhinus built on a

linear magnification basis of x 5. Roof of snout damaged. Photograph shows model reduced

to x 2 original size of skull.

of the nostril just behind the internarial process. It appears to be very similar

to the canal described by Brink (19616). Similar canals are present in the higher

cynodonts as well as in primitive forms like the scaloposaurids.

In medial view, the palatal process of the premaxillary is bifid. The upper

prong (Figs 1, 3A, 4B, 5, UPPPPX) is long and extends backward as far

as the canines (Figs 1, 3A). The shorter lower prong (PPPXSP) in dorsal

THE CRANIAL MORPHOLOGY OF THRINAXO DON LIORHINUS SEELEY 343

Fig. 3. A. Transverse section showing relationship of vomer, premaxillary and maxillary, and

replaced upper canine; B. Transverse section showing replacing canine. Both x 3.

ANNALS OF THE SOUTH AFRICAN MUSEUM

“Wy A.

A %,

ZN Gs

gas

eqs

welling,

ax “Wy B.

NY “yy,

Vy,

MX.==3 ( 5 a N.

DV,

FIN D UPPPPX,

. cu PX.

Sa PPPXSP

Ul3.

EC:

ROLI3

Fig. 4. Transverse sections through snout

A. Septomaxillary and palatal process of

premaxillary. B. Double anterior end of vomer

and palatal process of premaxillary. x 3.

ae INPPX.

: UPPPPX.

MPPX.

FNNG?—— if PPPXSP.

SPBPPPX.

Fig. 5. Graphic reconstruction of medial view of premaxillary. x 5

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 345

view is partly hidden by the upper prong (Fig. 4B). In palatal view (Fig. 1), how-

ever, it largely obscures the upper (dorsal) prong. Ventrally at the base of the

palatal process is a very small process (Figs 1, 4A, 5, SPBPPPX) situated on a

low ridge along the medial ventral margin of the palatal process. A similar

process is present in the cynognathids and Diademodon (Brink 1955a).

The lateral part of the premaxillary (i.e. palatal process excluded) forms the

anterior wall of the pit for the lower canine. The premaxillary extends backwards

medially alongside the ventral edge of the pit. This extension of the premaxillary

(MPPX), referred to as the maxillary process of the premaxillary, is held in a

groove along that part of the maxillary which forms the medial wall of this pit

for the lower canine (Fig. 3A). A similar relationship is found in Diademodon

and the cynognathids.

The anterior palatal foramen or foramen incisivum (Figs 1, 4A, B, FIN)

is enclosed between the maxillary and palatal processes of the premaxillary.

The ventral surface of the anterior part of the vomer (V) is clasped between the

two palatal processes of the premaxillaries (Figs 1, 4A, B).

Maxillary (Figs 1-14A)

Most of the secondary palate is formed by the maxillary. The posterior

palatal foramen (Figs 1, 9A, PPF) is about 2 mm long. It is situated on the suture

between the maxillary and the palatine, but is largely formed by the maxillary.

On the side wall of the snout the dorsal region of the maxillary (MX) lies

lateral to the nasal (Figs 2, 3, 6B) and its posterior region overlaps the lacrimal

(Riss, 2, 6B, 9B, L).

It is interesting to note that many foramina (see e.g. Figs 8A, B, RHC),

both large and small, perforate the anterior regions of the dentary, premaxillary

and maxillary. The canals from these foramina join a single large canal in each

side wall of the snout and in each ramus of the lower jaw. This fact lends

support to the suggestion originally made by Watson (1931) when discussing

the maxillary of the bauriamorph Ericiolacerta parva. He stated (p. 1168): “This

surface is perforated by many foramina. .. . These foramina must have trans-

mitted the blood-vessels and nerves to the skin of the face, their size suggesting

that the nerves were exceptionally big, a condition only to be explained by the

existence of a series of important tactile sense-organs grouped around the end

of the muzzle, that is of what ina mammal would be a rhinarium.’ However,

Estes (1961) has shown that the vascularization of the Thrinaxodon snout does -

not necessarily indicate a rhinarium, since Van Valen (1960) has described

similar vascularization in living reptiles such as Tupinambis. Estes claimed that

the numerous foramina actually preclude the presence of a movable muscular

cheek and lip.

Unfortunately, preservation of the anterior part of the maxillary is not perfect.

There appears to be little doubt, however, that a big foramen is situated on the

anterior part of the maxillary which forms the wall of the pit for the lower

canine. (Estes (1961, pl. 2, fig. 1) actually shows a large foramen in this position.)

346 ANNALS OF THE SOUTH AFRICAN MUSEUM

IPAR.

SUOC

IN == ill

Ow ‘i Ml

0. 7°

EPT. GxpaLvi. BS.WST. PRO. opo.

sectioned in median-sagittal plane. B. Lateral view; quadrate and quadratojugal removed. x 2.

py AWELC.

SPBPPPX.

Fig. 6. Graphic reconstructions of serially sectioned skull. A. Median-sagittal view. Hatched surfaces indicate structures

This foramen may be the anterior end of the canal probably for the main branch

of the maxillary nerve (see e.g. Figs 9, 10, CMN) which receives many smaller

branches from other parts of the snout.

In sections it appears as if the canal runs back into the base of the canine

alveolus, behind which it re-emerges as a distinct canal in which it is assumed

the maxillary ramus of the trigeminal nerve and blood-vessels ran (Fig. 10,

CMN). Immediately behind the canine alveolus several canals open on to the

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 347

CONTACT

WITH N

MXPS

CONTACT

WITH L

RUPCo4s@7!

Fig. 7. Graphic reconstruction of maxillary in medial view.

SeNANS

Fig. 8. Transverse sections through snout and

lower jaw showing root of functional upper

canine, foramina and canals in maxillary and

dentary. Note splenial part of lower jaw sym-

physis in A. x 2,25.

348 ANNALS OF THE SOUTH AFRICAN MUSEUM

EPCr:

Fig. 9. Transverse sections through snout. A. Posterior

palatal foramen; B. Maxillary-palatine suture in this

region. x 2,5.

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 349

outer surface of the maxillary. These canals join the maxillary canal which

runs backwards in the bone. A small canal joins the base of the first postcanine

alveolus with the maxillary canal running above it (Fig. 10). Further back the

maxillary canal makes direct contact with the bases of the second and third

postcanine alveoli. Several more little canals penetrate the maxillary walls of the

postcanine alveoli and open into the alveoli. From here the nerves and/or blood-

vessels in the canals may have been joined to the nerve and blood-vessels in the

maxillary canal. After entering into the base of the third alveolus the main

Mx;

| aan i eo

= Stila a,

Be Cs at

Fig. 10. Graphic reconstruction of canal system in wall of snout. x 2,5.

maxillary canal is joined by a second one, the external opening of which is in

the alveolar wall of the third postcanine (Figs 10, 13B, (F)DTCMN). These

two canals enter (Fig. 13B,, FDTCMN) into the ‘maxillary antrum’ (Fig. 13Bs,

MAN) above the fourth postcanine alveolus and at a level just behind the trans-

verse part of the maxillary-palatine suture. The foramen for this second branch

of the ramus maxillaris of the trigeminal is also shown by Estes (1961, pl. 2,

fig. 1). Kiihne (1956, figs 4a, 5a) found that in Oligokyphus the ramus maxillaris

also splits into two branches.

The maxillary antrum is formed in mammals by the maxillary only. In

Thrinaxodon the posterior part of the maxillary has a considerable overlap with

the lower part of the lacrimal laterally (see Figs 6, 10, MX, L) to form the outer -

wall of a cavity (Figs 11A, 13B,, MAN). The floor of this cavity is formed by the

maxillary, and its medial wall is formed by a dorso-laterally extending lamina

of the palatine (Figs 12, 13B,, P). This palatine lamina overlaps the lower part

of the lacrimal, which forms the roof of the cavity, medially. The ‘maxillary

antrum’ thus formed extends back as far as the sixth postcanine. Its medial wall

is not complete, being interrupted behind the palatine lamina where the antrum

is only a high, but shallow recess below the anterior part of the lacrimal. The

palatine and lacrimal are then reunited in a serrated suture to form a solid

eS)

50 ANNALS OF THE SOUTH AFRICAN MUSEUM

CMDN.

EDLBMN. |

FRAC IN ilies

Ne NTT NT

LF.

Fig. 11. Transverse sections through snout in the region of: A. Choanae and

maxillary antrum; B. Roof of n

Note: relationships of maxillary, palatine, lacrimal a

structure of lower jaw especially as far as position o

cerned. x 3.

asopharyngeal passage and lacrimal foramina.

nd jugal; shape of vomer,

f mandibular canal is con-

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY a5,

(CSEE FIGI3B] By) B2 B3

PRE

Fig. 12. Medial view of wall of snout of enlarged

wax reconstruction of serially sectioned skull

to show shapes and relationships of bones

along the various parts of the maxillary

antrum. x 2,5.

bridge of bone lateral to which the ‘maxillary antrum’ continues (Figs 11A, 12).

At this level the jugal (Fig. 11B, J) takes part in the lateral wall of the antrum.

It is held between the lacrimal medially and a lamina of the maxillary laterally.

The anterior palatine-lacrimal contact is 1,2 mm and the posterior bridge 1,4 mm

in length. Beyond this bridge the antrum continues as a low, but deep recess

2,6 mm long, between the lacrimal and the maxillary (Fig. 12). The recess

becomes progressively lower and ends at the level of the posterior margin of the

sixth upper postcanine.

In lateral view the foramen of a canal (Figs 10, 13B, IMC), which pene-

trates the maxillary for 3 mm, is present at the level of the anterior border of the

fourth upper postcanine and the ‘maxillary antrum’. This canal opens posteriorly

into the ‘maxillary antrum’ at the level of the palatine-lacrimal bridge. Its actual

opening (Figs 10, 11B, EIMC) is medially at the base of the lateral lamina of the

maxillary and directly below the lateral lower margin of the lacrimal, both of

which take part in the formation of the outer wall of the antrum. The signifi-

cance of the canal described here is uncertain, but it may have been for a third

branch of the maxillary ramus of the trigeminal nerve. The latter appears to

have swung down from behind the ‘maxillary antrum’ on to the dorsal surface

of the palatine. Here it might possibly have been joined by the ramus palatinus

of the facial nerve (Fig. 16A, CRPAL,,;;) which probably left the palate through -

the posterior palatal foramen (Figs 1, 16A, B, PPF). Both branches then appa-

rently continued further backwards on the pterygoid. Their further course is

discussed below (see Pterygoid, p. 357).

Lacrimal (Figs 2, 6, 10-14A)

The lacrimal forms the anterior wall of the orbit (Fig. 6B). It extends for-

ward as far as the third upper postcanine and the transverse part of the maxillary-

palatine suture on the secondary palate. In sections the lacrimal appears mainly

Sa ANNALS OF THE SOUTH AFRICAN MUSEUM

CONTACT a

WITH P CONTACT

WITH PT.

Bo. B3.

LD. L.

FLBMN®?

IMC:

FRAC.

UPC,.

Fig. 13. A. Graphic reconstruction of medial view of lacrimal.

B. Canals in wall of snout also shown in Fig. 10. Numbered levels

coincide with those shown in Fig. 12. B, and B; are transverse

sections from levels numbered 2 and 3 in B, or B, and B; in Fig. 12.

<2:

as a thin sliver of bone with the exception of a thickened middle part which

contains the lacrimal duct. In lateral view the lower part of the bone is over-

lapped by the posterior part of the maxillary, and the dorsal part of the lacrimal

is just overlapped by the ventral margin of the posterior part of the nasal

(Fig. 6A, MX, L, N). Unfortunately the preservation of this part of the snout

is not as good as could be desired and the sections are difficult to interpret. The

lacrimal also overlaps the anterior ventral part of the prefrontal (Fig. 6A, PRF).

The anterior opening of the lacrimal duct (Figs 6A, 10, 12, 13A, B, ELD) is

near the anterior end of the bone just above and in front of the anterior part

of the wall of the ‘maxillary antrum’. The maxillary forms a distinct ledge below

the opening. In a posterior direction this ledge becomes more pronounced to

form the floor of the ‘maxillary antrum’, but in front of the opening the ledge

very gradually becomes indistinct.

Where the palatine and the lacrimal form the bridge which is part of the

inner wall of the ‘maxillary antrum’ described above (see Maxillary), the lower

THE CRANIAL MORPHOLOGY OF THRINA XODON LIORHINUS SEELEY 353

part of the lacrimal is wider than anywhere else when seen in section (see e.g.

Fig. 11B, L). Between the fourth and fifth upper postcanines is the anterior

foramen (Figs 10, 13B,, FLBMN) (Estes (1961, pl. 2, fig 1) also shows this

foramen) of a canal (Figs 10, 13B, DLBMN) which runs in the lacrimal just

below the lacrimal duct. The lacrimal duct itself divides into a small upper and a

larger lower duct only | mm behind the level of the foramen just mentioned

(Figs 10, 13B) and opens (Figs 10, 11B, 13B, LF) into the orbit only 0,6 mm

behind the division. The other canal turns sharply downward (Figs 10, 11B, 13B

DLBMN) at this level and enters (Fig. 11B, EDLBMN) into the posterior part

of the ‘maxillary antrum’, the inner wall of which is formed at this level by the

palatine-lacrimal ‘bridge’. This canal appears to be comparable with the one

which Kiihne (1956: 28, fig. 5) described in Oligokyphus as the canal for the

lacrimal branch of the maxillary ramus of the trigeminal nerve.

It seems reasonable to assume, on the basis of the canals described above in

the maxillary and lacrimal, that the ramus maxillaris of the trigeminal nerve

was made up of the following branches (Fig. 10):

(a) the main anterior branch made up of smaller branches from the tip of the snout,

the incisors, the canine, the first two postcanines and the snout in their vicinity,

joined by

(6) asecond branch from the side of the snout in the vicinity of the third postcanine.

These two branches (a and 5) entered the ‘maxillary antrum’, in the posterior part

of which they were joined by

ramus maxillaris, and

(d) a lacrimal branch.

The compound maxillary ramus then left the ‘maxillary antrum’ and might have

been joined a little further back by the ramus palatinus of the facial nerve.

The wide basal part of the lacrimal just in front of the narrower posterior

end of the bone has a somewhat spongy structure which makes it difficult to

reconstruct the course of a possible canal in this part of the bone from serial

sections. There does, however, appear to be a canal which may then be compared

with the vascular canal found by Kiihne (1956: 28, fig. 5) in Oligokyphus.

Palatine (Figs 1, 6A, 9B, 11A—-16B)

foramen, but it is overlapped on its medio-dorsal surface by the palatine (Figs

11A, 12, 13B,, 13B;, P, MX). This overlapping dorsal part of the palatine has a

very thin upper rim of which the margin is expanded laterally and medially to

form a distinct ridge. It is this thin upper end which forms the medial wall of

the anterior part of the ‘maxillary antrum’ (Figs 12, 13B,, P, MAN), the ridge

being in contact with the medial surface of the lacrimal. The palatine and the

maxillary are joined by a thin ledge which projects from the medial border of the

maxillary ito a groove along the lower lateral surface of the palatine (Figs 11,

354 ANNALS OF THE SOUTH AFRICAN MUSEUM

N 5 A.

PRE.

\ e

: ZN J

KVPG- (eS Girw~ TW, Te

LDUPG- wy WS nee ees

—— UPC.

a eee

COR,

( PA,

CMDN. : SS AN.

PRE.

F. B.

Ke:

as gp PT.

SS Shy ff | Sis EPTR.

VP.

UELPCo:

CMDN. SS)

Fig. 14. Transverse sections through skull and lower jaw; A. At level in anterior part of orbit;

B. Slightly further back. Note: anterior ends of pterygoids and ectopterygoids in B and of the

angulars in A; posterior ends of nasals and maxillaries in A and of the lacrimals in B; extension

laterally of jugals. x 2,5.

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 355

13B,, B;) (a schindylesis or wedge-and-groove suture). The two bones lie close

together at all their contact surfaces.

Ectopterygoid (Figs 1, 14A, 15A, 16A)

Parrington (1946) describes two foramina in this bone, but it seems to be a

variable feature since the sectioned skull shows only one (Fig. 1, FEPTR).

This single foramen is situated almost in the position of Parrington’s anterior

one, but it lies nearer the inner surface of the ridge.

Vomer (Figs 1, 3, 4, 6A, 8A-9B, 11A—12, 13B., 14A—-15A)

The vomer has several features of interest. The posterior curved plate-like

part forms the roof of the nasopharyngeal passage behind the secondary palate

(Figs 1, 11B, 14A, B, (K)VP(G)). Antero-dorsally it is overlapped by the

palatines (Figs 1, 4A, P) and postero-dorsally by the pterygoids (Figs 1, 14B,

PT). The part comprising the thin median vertical plate extends forward into the

lower half of the snout (Figs 1, 6A, 11A, V) to be clasped anteriorly by the upper

prongs of the palatal processes of the premaxillaries (Figs 1, 3A, 4B, 6A,

UPPPPX). Nowhere is the median vertical plate of the vomer in contact with the

maxillaries. The front end of the vertical plate, which extends forward to a level

with the fourth incisors, is bilaterally paired for a distance of 3,4 mm (Figs

1, 4B, 6A). A line of fusion can, however, be distinguished much further back.

Anteriorly the lower end of each plate curls slightly outwards, but behind the

line of fusion and as far back as the posterior border of the lower canine pit

they curl pronouncedly (Figs 1, 3A, B). This condition is also found in higher

cynodonts investigated by the present author and by Broili & Schréder (1934a).

It need not cast doubt on the homology (proposed by Parrington & Westoll

1940) of this structure and the mammalian vomer which usually arises singly,

since De Beer (1937: 434) gives examples of a paired origin of the vomer in

mammals.

Together with the upper prongs of the palatal processes of the premaxillaries

(UPPPPX) which also extend back to the level of the posterior margins of the

lower canine pits, the vomer with its curling lower ends forms a rod in which the

component parts support each other. This rod, of which the ventral surface is

grooved (due to the curling vomer), divides the large foramen, situated ventrally

between the premaxillaries, into a left and right foramen incisivum (Figs 1, 4A, B,

FIN). The palatal plates of the maxillaries form the posterior border of each.

foramen. Above the palatal plates of the maxillaries, but not in contact with

them, lies the vertical vomer plate of which the ventral margin is, in this region,

thickened and rounded (Figs 1, 8A, B, V). The upper margin of this vomer plate

has a distinct groove (Figs 3, 6A, GR). The groove extends back over the whole

length of the vomer (Figs 3, 6A-11B, 14A), presumably over the pterygoids

(where it may, however, be interrupted) and on to the rostrum of the para-

sphenoid (Figs 6A, 15C, 16A, 17, PCP). A similar condition is found in the

cynognathids and in Diademodon (Brink 1955a). This groove probably supported

356

ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 15. Transverse sections through skull and lower jaw in

orbital region. Note: A. Posterior ends of vomer and ecto-

pterygoids and anterior ends of postorbitals; B. Posterior

ends of prefrontals and palatines and anterior end of rostrum

of parasphenoid; C. Relationship between pterygoids and

parasphenoid rostrum behind transverse processes of ptery-

goids. A x 1,9. B-C x 1,75.

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 357

a cartilaginous internasal and interorbital septum. In the higher cynodonts the

interorbital septum is apparently replaced by a presphenoid.

The ventral palatal extensions of the premaxillaries, maxillaries and pala-

tines form a secondary palate. The vertical plate of the vomer lies above the

median longitudinal suture of the palate but is not in contact with it. The two

halves of the palate meet each other in a sutura harmonia. Along the median

dorsal margin of each half of the palate runs a distinct ridge (see Figs 8B, 9).

The secondary palate is incipient in Gorgonopsia, where the anterior end of the

vomer is broad and flat.

In Thrinaxodon and higher cynodonts the development of the secondary

palate is much more advanced than that of Gorgonopsia. The transformation of

the anterior end of the vomer from a broad flat structure to a vertically orientated

sheet of bone and the development of a secondary palate, without which breath-

ing would be difficult when food was chewed, are probably related to the develop-

ment of homoiothermy. The secondary palate was probably necessary for a

proper chewing process, which in turn was probably a prerequisite for quicker

liberation of energy from the food taken in. A larger supply of energy was

apparently needed in the more active advanced mammal-like reptiles. The

enlarged nasal passage resulting from the development of the secondary palate

was probably utilized for expansion of the surface of the mucous membrane,

since the turbinals, for the presence of which there are indications (see p. 383),

were probably rudimentary. Such an increased surface of mucous membrane

would have been necessary to moisten and warm the increased supply of air

taken in due to increased activity in those animals with increased metabolic

rate and incipient homoiothermy.

Brink (1957) discussed the question of homoiothermy and related develop-

ments in the advanced therapsids. The nasal passage itself was possibly com-

pletely divided (but the nasopharyngeal passage only incompletely) by the

vertical plate of the vomer, the probable internasal septum which appears to

have rested in the groove on the dorsal edge of the vertical plate of the vomer,

and an indeterminate structure between the lower end of the vomer and the

palate. Possibly the ridge along the median dorsal edge of each half of the palate

and the thickened lower margin of the vertical plate of the vomer indicate the

attachment areas of a strong connective tissue connection between these

structures.

Pterygoid (Figs 1, 6, 14B-17C)

Sectioning has revealed a considerable amount of additional information

about the pterygoids. The antero-medial portion of each pterygoid is overlapped

ventrally by the horizontal plate of the vomer (Figs 1, 14B, VP, KVP, PT),

and the antero-lateral portion is overlapped dorsally by the palatine (Figs 14B,

I5A, P, PT). It may be noted here that the structures indicated by Broom

(1938a, fig. 9) as pterygoid and maxillary, are really the palatine and pterygoid

respectively.

358 ANNALS OF THE SOUTH AFRICAN MUSEUM

EPTR

PT.

FEPTR

FPT.

CRPALVII?

CRMAX V?

Fig. 16. Graphic reconstructions of: A. Part of palate and basicranium in dorsal view;

B. Medial view of palatine; C. Medial view of pterygoid. x 2,5.

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 359

The relationships of the pterygoids with the other basicranial elements

behind the transverse processes of the pterygoids which guide the lower jaw

(Figs 15B, 16C, TPPT) are of considerable interest because of the nature of their

sutures with the parasphenoid and epipterygoids. The rostrum of the parasphe-

noid (Figs 15B, C, 16A, 17, PCP) lies between the posterior extensions of the

pterygoids. These extensions have interdigitated sutures with each other and

with the rostrum (Figs 15, 17, PT, PCP). Fractures reveal a similar condition in

cynognathid skulls investigated by the present author. At the level where the

anterior part of the rostrum appears in ventral view, each pterygoid divides

miorwe, bars(Fic. 17A, IDIOPPT; Figs I, 16A, C, 17B, C, IPPT, OPPT).

Estes (1961) indicates very distinct interpterygoid vacuities at this level

in juvenile Thrinaxodon skulls and suggests that in older specimens these vacuities

are obscured by appression caused by their stronger pterygoid musculature. It is

possible that the vacuities described by Estes are not natural structures, since

he states that both the juvenile skulls investigated by him were crushed dorso-

ventrally, which may have been responsible for the formation of ‘interpterygoid

vacuities’.

The inner bars (IPPT) which are joined to the rostrum by interdigitated

sutures (Fig. 17B) extend back for 5 mm (Figs 1, 16C). Broom (1938, fig. 12)

mistook these inner bars for the epipterygoids. In a posterior direction the inner

bars taper to points which fit into shallow pits in the rostrum of the para-

sphenoid (Fig. 17C) exactly as described by Olson (1944). About 1,5 mm behind

the division of the pterygoid, the outer bar (OPPT) widens outwards and

upwards to support the anterior ventral tip of the epipterygoid (Figs 1, 16A,

17B, EPT) which rests in a groove on this outer pterygoid bar. In section this

tip of the epipterygoid rapidly expands in a posterior direction to form a distinct

antero-ventral process of the epipterygoid (see Fig. 22). This part of the epiptery-

goid appears to agree closely with the same structure in Oligokyphus (Kuhne

502555 pl: 10; fig. 1b).

The widened outer pterygoid bar divides into two about 5 mm behind the

initial division of the pterygoid into an inner and an outer bar. The quadrate

ramus of the pterygoid (QRPT) separates from the dorso-lateral part of the

outer pterygoid bar (Figs 1, 16A, C, 17C). The remaining median part of the

outer pterygoid bar extends back only | mm behind the division before it meets

the basisphenoid in a serrated suture (Fig. 1, PTPSS).

Parrington (1946) describes ridges on the ventral surfaces of the pterygoids -

which continue on to the basisphenoid. The present investigation shows that a

ridge extends backwards from the posterior part of each palatine, on to the

pterygoid and finally on to the basisphenoid. The anterior palatine-pterygoid

part of the ridge (Fig. 1, PPRASP) has been described as a ridge for the attach-

ment of the soft palate, an identification which should probably be extended to

include the more posterior part of the ridge formed by the pterygoid only

(PTRASP), but this interpretation results in the assumed presence of a very

long, soft palate. The pterygoid part of each ridge is formed by the median

360 ANNALS OF THE SOUTH AFRICAN MUSEUM

PAR,

F A.

po, \

Fig. 17. Transverse sections through skull and lower jaw showing relation-

ships: between frontals, parietals and postorbitals; between pterygoids,

epipterygoids and parasphenoid rostrum; between elements of lower jaw.

AX 1,8..B Kove tie eS:

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY

PAR. PF. A.

———— GRVSC.

DF. D CROPy

BPTPL sie

Ga=: QRPT.

: BSCROPPT.

PCP. GRPR Vil,

EPT.

DF CROP

GRMAXy

QRPT. BPTPL

BS, PTP.

BSCROPPT,

GRPR VI.

di, A

an «a

AN.

PA,

Fig. 18. Transverse sections through braincase and lower jaw showing

parietal foramen, relationships between parietals, epipterygoids and basi-

sphenoid and between carotid canals and sella turcica. Note surangular,

angular and prearticular relationships. x 1,5.

361

362 ANNALS OF THE SOUTH AFRICAN MUSEUM

ventral part of the outer pterygoid bar and its posterior end by a ridge on the

basisphenoid (Figs 18A, B, C, BSCROPPT) just lateral to the keeled posterior

part of the parasphenoid rostrum (see Fig. 1). The basisphenoid part of the

ridge forms the median ventral margin of the basipterygoid process (Fig. 1,

BPTP).

Between the keeled posterior part of the parasphenoid rostrum (Parring-

ton’s (1946) ‘ridge on the processus cultriformis’) and the closely applied median

ventral part of the outer pterygoid bars, deep grooves are formed. These grooves

continue backwards on to the ventral surface of the basisphenoid (Figs 1, 18A, B,

GRPR,,;) as far as the internal carotid foramina (Fig. 1, ICF). Olson (1944)

suggested that the rami palatini of the facial nerves passed forward in these

grooves and then turned upwards between the pterygoids and basisphenoid,

anterior to the basipterygoid processes. Parrington (1946) questioned this

suggestion because he thought that Olson did not show clearly where the nerve

re-entered the skull. This uncertainty apparently rests on a misunderstanding of

Olson’s description. The point of re-entry could lie in the fork between the inner

pterygoid bar (suturally joined to the rostrum) and the outer pterygoid bar

(Figs 1, 16A) in what may be, according to the description of Estes (1961), the

remnant of the interpterygoid vacuity. The canal is difficult to see without

sectioning, because the ventral parts of the outer pterygoid bars are inclined

towards each other (Fig. 17B, C, OPPT). A similar condition exists in the cyno-

gnathids which were investigated. In these forms the probable course of the

ramus palatinus of the facial nerve can easily be followed forward from its

exit from the braincase.

Parrington (1946) suggested that a small foramen which he found only in

one specimen on the left side between the basipterygoid process and the ptery-

goid might have transmitted the ramus palatinus. The actual relationships

between the bones, as determined by sectioning, differ somewhat from those

figured by Parrington, and the foramen probably represents a small opening in

the suture between the two elements.

Probably the ramus palatinus of the facial nerve passed through the ptery-

goid-parasphenoid canal described above (Fig. 16A, CRPAL,;;). From here

it may have passed backwards lateral to the internal carotid foramina, on to the

upper half of the lateral face of the basisphenoid. In the region of the unossified

zone it may have passed on to the concave lateral face of the prootic on which it

probably ran gradually upwards. It probably passed below the lateral lamina of

the prootic (PROF) and entered its foramen which is situated just below and

behind the large foramen for the trigeminal nerve. The maxillary ramus (Figs |,

16A, CRMAX,) of the trigeminal nerve after leaving the ‘maxillary antrum’

probably continued in the groove on the dorsal surface of the pterygoid until it

penetrated the foramen pterygoideum (FPT). From here it probably continued

along the latero-ventral surface of the outer prong of the pterygoid (OPPT)

until the latter’s quadrate ramus (QRPT) separated and the antero-ventral

process of the epipterygoid became laterally exposed. From this point the

ee gato,

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 363

PAO.

QREPT.

SAN.

PAR,

aL | =

Q. = BO: ®

i ian peer

AN Sg PA. AS________— san.

PAR,

Fig. 19. Transverse sections through braincase and lower jaw. A. Trigeminal

foramen, tips of pila antotica and antero-dorsal process of prootic and unossified

zone in basicranium. B. Suspensorial region and lateral lamina of prootic.

C. Retroarticular process of articular, anterior vertical semicircular canal of

internal ear, and relationships between lateral lamina, quadrate and stapes.

AG ET, BC << ¢. 15.

364 ANNALS OF THE SOUTH AFRICAN MUSEUM

maxillary ramus probably continued along the groove (Fig. 18, CRMAX,) on

its surface, just above the level of the quadrate ramus of the pterygoid, until it

reached the ganglion Gasseri lying within the foramen (F,) between the epiptery-

goid and prootic (Figs 1, 16A). The abducent nerve probably followed much the

same course as the palatine ramus of the facial nerve and entered its foramen

(Fig. 6A, F,,) anterior to and below the foramen for the former. The geniculate

ganglion of the facial nerve (VII) probably lay in the posterior part of the fora-

men between the epipterygoid and the prootic, in a recess under the prootic.

After entering its foramen the palatine ramus of the facial nerve had only to

run directly upwards for a very short distance before joining its ganglion.

Basisphenoid and parasphenoid (Figs 1, 6, 15B-16A, 17A—19B, 20, 21C, 22B,

25A, 28, 29A)

Sectioning revealed no more information than that given by Olson (1944)

and Parrington (1946), and that given above under the section on the pterygoid.

An unossified zone is present between the basisphenoid and basioccipital

(Figs 1, 6A, 16A, 19A, 20, UZ). The parasphenoid is well developed laterally.

Earlier preparation of the sectioned skull exposed the basicranium so that it is

not possible to be sure exactly how far the parasphenoid extends backwards on

the ventral surface. It does not cover the unossified zone and the anterior part

of the basioccipital, but this may be due either to damage or to the fact that the

specimen is fairly young (Fig. 1, WPS).

PAR.

Fig. 20. Transverse section through braincase and lower jaw showing rela-

tionship between epipterygoid and prootic, abducent foramen (see also

Figs 21C and 24A), and unossified zone in basicranium. x 1,5.

In ventral view the wings of the parasphenoid do not extend as far laterally

towards the fenestra ovalis as they are figured by Parrington (1946). Again the

difference may be ascribed to different stages of growth, since Estes (1961)

also figures the posterior ends of the wings of the parasphenoid in very young

specimens as lying comparatively farther forward than shown by Parrington.

The dorsum sellae is practically non-existent and therefore the sella turcica

(SET) is very shallow and narrow and flanked laterally by thin bone walls

THE CRANIAL MORPHOLOGY OF THRINAXO DON LIORHINUS SEELEY 365

(Figs 16A, 21C, SET, WST). The internal carotid canals emerge almost directly

above their points of entry into the parasphenoid-basisphenoid (Figs 1, 16A,

ICE 18C; (CCA).

Fig. 21. A. Ventral view of part of basicranium (after

Parrington 1946) showing pila antotica and basiptery-

goid process. B. Ventral view of basicranium of a scalo-

posaurid (after Crompton 1955) for comparison with

A. C. Postero-dorsal view into braincase of enlarged

wax reconstruction of serially sectioned skull showing

shallow sella turcica, basicranial process, abducent

foramen and pila antotica. C x 2,5.

The parasphenoidal teeth described in juvenile Thrinaxodon specimens by

Estes (1961) may be similar to the irregular surface of the thickened medio-

ventral margin of the maxillary and palatine in cynodonts. The median bony

ridge thus formed appears comparable with the raphe palati of the mammalian

palate. Seeley (1908) described the irregular surface of the maxillary and palatine

as a dental armature and Watson (1911) referred to granules with enamelled

tips. Broom (1911) and Broili & Schréder (1934a) rejected these descriptions.

The present author regards the irregular surface of the median ridge as part of

the attachment for a mucous membrane which had to be firmly fixed so that

forceful movements of the tongue would not dislodge it. In mammals the lamina

propria of the mucous membrane is usually continuous with the periosteum of

the bone above. Laterally the mucous membrane is not evenly adherent to the

bone and is connected to it by strong bundles of connective tissue. The epithe-

lium of the hard palate, which has to be able to withstand wear and tear, is

366 ANNALS OF THE SOUTH AFRICAN MUSEUM

attached to the median ‘raphe’ by a thin lamina propria. Rugae with connective

tissue cores radiate laterally from the raphe, the uneven surface of which may

serve to aid in the attachment of the connective tissue cores (to the periosteum).

A small foramen, described by Estes (1961) and confirmed by the current

investigation, lies in each wing of the parasphenoid just medial to the fenestra

ovalis and anterior to the opisthotic contact with the basioccipital (Figs 1, 25A,

FN,). This foramen is situated in the precise position where the mammalian

Eustachian tube opens.

ae ANREL.

ADP. PRON

PRO.

Poo, ke

T. : —= >

ak <BPTPL . Veer \ Les Reser,

A. B. D. F. G.

Fig. 22. A. Graphic reconstruction of epipterygoid and prootic

in lateral view to show probable system of veins, contact with

frontal, and levels of sections given in C. B. Transverse section

showing frontal-epipterygoid contact, detaching quadrate ramus of

pterygoid and postdentary bones of lower jaw. C. Sections indi-

eaied in AVA X 3: B xXIES. C XS 2k:

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 367

Epipterygoid (Figs 1, 6A, B, 16A, 17B-19B, 20-22, 28)

The only feature of this bone which has not been described in earlier

publications is the relationship of the tip of its antero-ventral process to the

pterygoid, described above under Pterygoid (p. 359). The slight overlap of the

frontal and epipterygoid (Figs 1, 6A, B, 22A, B, C) is a feature also found in the

cynognathid skull and it is described by Crompton (1958) in Diarthrognathus

as well. In Thrinaxodon the anterior part of the lateral flange of the frontal is

laterally overlapped by the prefrontal (Figs 6, 15A, B, F, PRF). The anterior part

of the postorbital in turn partly overlaps the posterior end of the prefrontal, but

actually lies mainly dorsal to the bone (Fig. 6B, PO, PRF). Further back the

lateral flange of the frontal is laterally overlapped by the thin front end of the

parietal of which the lower part in turn is laterally overlapped by the posterior

end of the postorbital (Figs 6, 17, PAR, PO). Between the posterior end of the

prefrontal and the anterior end of the parietal only the postorbital overlaps the

frontal flange. Behind the postorbital only the thin parietal sheet overlaps the

frontal flange (Figs 6, 18A, B). At this level the frontal forms the anterior and

lateral lower margins of the wall of the parietal foramen (PF). In sections the

posterior lower part of the lateral flange of the frontal anterior to the parietal

foramen becomes progressively longer, until its lower end makes contact with

the anterior upper edge of the epipterygoid (Figs 6, 22, F, EPT). This ventrally

projecting part of the frontal ends well in front of the upper part of the bone and

the parietal foramen. In this region the lower end of the frontal flange protrudes

well below the parietal which overlaps its dorsal part (Fig. 6B).

Sectioning, therefore, shows that it is not just a slender horizontal process

of the frontal which meets the epipterygoid. The median wall of the orbit is formed

in the same way in the cynognathids and Diarthrognathus, and closely resembles

the condition in mammals and tritylodontids.

WATSON COX THRINAXODON

Fig. 23. Three diagrams to show various interpreta-

tions of venous system around braincase. x c. 2,5.

Prootic (Figs 1, 2, 6, 16A, 19A—22A, 22C-24A, 25A, 26, 27, 29A, B)

The anterior part of the prootic consists of the antero-dorsal process

(ossified taenia marginalis, Crompton 1955b) which forms the postero-dorsal

margin of the trigeminal foramen (Figs 22A, 24A, ADP), and the antero-

ventral process, identified by Parrington (1946) and Crompton (19555) as an

368 ANNALS OF THE SOUTH AFRICAN MUSEUM

ossification of the pila antotica (PAO) which forms the postero-ventral margin

of the trigeminal foramen. The postero-dorsal part of the epipterygoid, which

forms the antero-dorsal margin of the trigeminal foramen, projects on to the

anterior part of the antero-dorsal process to which it is joined by a serrated

suture. It is held between a large outer plate and a slender inner process (Figs

6A, 22A, 28, ADP). The notch described by Olson (1944: 16-17), dorsal to both

the prootic incisure and the latero-dorsal process, is closed off by the lower

margin of the parietal to form the foramen for the vena capitis dorsalis (Figs 6A,

22A, FSC) (see also pp. 373-5).

SC. —> coors

ONTACT.

WITH SUOC.

CONS T WITH go

RO. i>

FO,

CONTACT WITH BO.

SCN At

a CON SUS

Fig. 24. Graphic reconstructions of

medial views of: A. Prootic; B. Opistho-

tic; C. Squamosal) x, 2:5:

Some detail can be added to the descriptions by Olson (1944) and Parrington

(1946) of the foramen for the abducent nerve. The nerve passes through a deep

notch in the antero-ventral edge of the antero-ventral process (pila antotica)

(Figs 20, 21C, 24A, F,,). The notch is closed anteriorly by a postero-dorsally

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 369

directed flange of the lateral wall of the sella turcica which lies in the dorsal

surface of the basisphenoid (Figs 6A, 16A, 21C, BS, WST, F,;). Olson’s (1944)

figure is therefore more closely comparable with the structure as determined in

the present investigation than Parrington’s (1946) figure in which the foramen

lies entirely within the antero-ventral process.

Sectioning shows that the foramen lies behind the sella turcica and the

unossified zone. The foramen for the facial nerve (Figs 24A, 27, 28, F,,.) is quite

distinct immediately below and behind the prootic incisure (posterior wall of

the trigeminal foramen) from which it is separated by the prefacial commissure.

The ramus ophthalmicus profundus of the trigeminal nerve probably ran

forward along the medial surface of the lower third of the epipterygoid from

the Gasserian ganglion lying just within the trigeminal foramen. Its possible

course is indicated by a slight though definite groove on the medial surface of

the epipterygoid just above the level of the antero-ventral process of this bone.

On emerging from the braincase through the lower end of the foramen lacerum

anterium (Fig. 6A, FLA), the ramus ophthalmicus profundus would have been

immediately above the anterior ventral part of the epipterygoid (Fig. 18, CROP,)

which is in contact anteriorly with the pterygoid, and more posteriorly with

the basisphenoid (Fig. 18B, BPTPL).

Fig. 25. A. Antero-ventral view of posterior part of enlarged

wax reproduction of serially sectioned skull to show contacts

of stapes and quadrate. Note retroarticular process of articu-

lar. B1, 2, 3, 4. Lateral, ventral, anterior and medial views of

right stapes. Note positions of anterior, dorsal and posterior

planes in the different views. x 2,5.

370 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 26. Stereophotograph of anterior view of right posterior part of model of serially sectioned

skull showing squamosal recess with quadrate and quadratojugal, as well as outer surface of

prootic with grooves and canals for system of veins. Approximately x 3. (See also Fig. 22A.)

See accompanying diagram for lettering.

Fig. 27. Stereophotographs of: A. Inner surface of wall of right side of bra‘ncase; B. Floor and

ventrally situated structural features (jugular foramen, sacculo-cochlear recess and fenestra

ovalis). Approximately x 2,5. See accompanying explanatory diagrams for lettering. (See also

Figs 24A, 28.)

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 371

Fig. 26. Explanatory diagram.

Fig. 27. Explanatory diagram.

372 ANNALS OF THE SOUTH AFRICAN MUSEUM

The paroccipital process is formed by a postero-lateral flange of the prootic

(see Fig. 26) which rests dorsally on the opisthotic in a squamous suture (Fig.

29B, PRO, POP, OPO). It abuts against the squamosal with only its dorsal and

ventral margins, but possibly had a cartilaginous cap which became ossified

during later stages of growth than that represented by the sectioned skull.

The step-like suture described by Parrington, between the prootic and opisthotic

(Figs 1, 25A, POP, PRO) which forms part of the roof and the hind wall of the

middle ear cavity, is confirmed. The prootic flange (lateral lamina) (PROF)

forms the anterior border of the pterygo-paroccipital foramen (Fig. 2, PTPOF)

and meets the quadrate rami of the pterygoid and epipterygoid (Figs 1, 2, 16A,

19B, 22A, QREPT). The lateral lamina of the prootic and the quadrate ramus of

the pterygoid extend to the median margin of the quadrate (Figs 1, 17C, 26, Q).

PAR, SSeS a IPAR.

ce:

EXO,

PVC.

2 A ORSCRFJ.

: fa — Pamenaricenl ORSUSR.

SAC. BOE oS PANE.

FN ic rs > sl LLF xjI.-

BS ars

F Vi ——ar Se ais

CO: =). = So:

Q: ee —_ OPO.

Fig. 28. Graphic reconstruction of inner ear structures in relation

to the otic and occipital bones. x 4.

The subarcuate fossa is very distinct (Figs 24A, 27, 28, FSA). In an oblique

dorsal view backwards into the braincase this excavation for the lateral lobe of

the cerebellum appears to extend into a large space above the anterior opening

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 313

of the posttemporal fossa. There appears to be no vascular duct which pene-

trates the cranial wall from the subarcuate fossa as there is in Oligokyphus

(Kiihne 1956). (See, however, under Supraoccipital, p. 378.) A convexly curved

horizontal lamina separated the excavation for the lateral lobe of the cerebellum

from the horizontal canal and the vestibule (Fig. 27A). This lamina is also

described by Brink (1955a, fig. 2b) for Diademodon, in which the conditions in

this part of the braincase agree very closely with those in Thrinaxodon. The rela-

tionship of the semicircular canals to the subarcuate fossa can be seen in

Figure 28.

The most interesting feature of the otic bones is the system of grooves,

canals, and foramina for veins which enter and leave the braincase and was

described by Watson (1920) and Parrington (1946). In a study of the dicynodont

Kingoria, Cox (1959) pays considerable attention to these veins. Since the system

of grooves, canals, and foramina in Kingoria appears to be closely comparable

with that of Thrinaxodon and of the cynognathids, Cox’s interpretations, with

those of the earlier investigators, are of considerable value for understanding

the conditions in Thrinaxodon. Cox suggests that the groove (Fig. 18, GRVSC)

along the junction of the parietal and epipterygoid of cynodonts represents the

course of the vena parietalis, and the foramen (Figs 6A, 22A, 28, FSC, 26, 27)

described by Watson (1911, 1920) between the parietal and prootic is the point

of entry into the braincase of the vena parietalis and the vena capitis dorsalis

(Watson’s ‘vein of sinus canal’) which, according to Cox, drained forward from

the occipital musculature. The course of this latter vein was through the post-

temporal fossa, then upwards into the posterior end of the dorsal groove (sinus

canal), and into the braincase through the above-mentioned foramen. The vena

capitis lateralis ran between the pterygo-paroccipital foramen (Figs 22A, 23,

PTPOF, 26) and the trigeminal foramen. The vena capitis dorsalis and the vena

capitis lateralis were joined by a vein which ran forward across the dorsal and

anterior surfaces of the paroccipital process from the posttemporal foramen

into the pterygo-paroccipital foramen, to provide an alternative method of

drainage. Cox’s interpretation is a modification of Watson’s (1920) and is

supposed to allow for the fact that the vein which occupied the posttemporal

fossa drained forward.

In Watson’s interpretation, which is made difficult because it gives a wrong

impression of the direction of flow of the blood, the vena capitis lateralis ran

through the pterygo-paroccipital foramen and was formed by three roots.

These were: one which ran forward through the posttemporal fossa; one which

ran backwards from the trigeminal foramen; and one which ran backwards in

and then downwards and forwards from the ‘sinus canal’ which is the groove

along the ventral parietal and dorsal prootic and epipterygoid borders. Watson

also described two smaller veins. The first is said to have left the braincase

through a foramen between the prootic and parietal and then to have joined a

vein which ran in the ‘sinus canal’. According to Cox this foramen is for the

vena parietalis and the vena capitis dorsalis. The second small vein described by

374 ANNALS OF THE SOUTH AFRICAN MUSEUM

Watson is said to have left the braincase through the upper part of the foramen

for the trigeminal nerve and then to have joined the third root of the vena

capitis lateralis which ran from this same foramen.

A study of the skull of Thrinaxodon reveals that the grooves and canals fit

the interpretation of Watson (1920) of the courses of the main veins as well as

they do that of Cox (1959) and that the correct interpretation may be a synthesis

of the two interpretations. Figure 26 shows a very distinct groove (VG) running

between the pterygo-paroccipital foramen and the trigeminal foramen. Both

Watson and Cox agree that a vein ran along this groove, Watson calling it a

root of the vena capitis lateralis, and Cox calling it the vena capitis lateralis.

The structure of the anterior dorsal face of the paroccipital process leaves no

doubt that a vein ran across it (see Fig. 26) between the posttemporal fossa and

the pterygo-paroccipital foramen. This vein must have been the second root of

the vena capitis lateralis described by Watson and which is regarded by Cox asa

communicating vein between the vein which he calls the “vena capitis dorsalis’

and the vein which he calls the ‘vena capitis lateralis’.

In order to understand the course of Cox’s vena parietalis and vena capitis

dorsalis, and Watson’s vein of the sinus canal, it is necessary to describe briefly

the relationship between the lower margin of the parietal and the upper margins

of the epipterygoid and prootic.

In the reconstructed skull of Thrinaxodon the parietal and the epipterygoid

do not make contact and the anterior part of Watson’s sinus canal is therefore

not distinct (Fig. 6A, EPT). Since the lower margin of the parietal is, however,

grooved (Fig. 18, GRVSC) it is possible that the epipterygoid and the anterior

part of the prootic are slightly dislocated from the parietal, or else that further

growth would have formed the groove or canal proper. The posterior part of

the groove (canal) is normal in the reconstructed skull where it is walled

medially by the anterior end of the supraoccipital (Figs 6A, 29A, SUOC).

Watson believed that a third root of the vena capitis lateralis ran along this

groove, but Cox believes that its anterior part was occupied by the vena parietalis

and its posterior part by the vena capitis dorsalis. The further course of the

vein which occupied the posterior part of the groove (or canal) can easily be

followed in the reconstructed skull where a distinct groove runs straight down-

wards from the horizontal groove, well in front of the eaves formed by the

squamosal over the posttemporal fossa (Figs 22, VCD, 26). It then continues

forwards and downwards on the paroccipital process at its proximal end where

it is continuous with the wall of the braincase. Watson’s description and figure

of the course of the third root of the vena capitis lateralis fit the groove described

above very well. Cox’s description of the upper part of the groove is also correct,

because at the base of the vertical part of the groove where it turns forwards

there is an equally distinct part of the groove which turns backwards (Figs 23

Thrinaxodon, 26) into the posttemporal fossa. This must be the course of the

posterior part of Cox’s vena capitis dorsalis. There is no doubt that the vein

which ran down the paroccipital process (see p. 373) splits from Cox’s vena

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 375

capitis dorsalis in the posttemporal fossa because their grooves merge into a

single large groove.

The reconstructed Thrinaxodon skull also distinctly shows that the third

root of the vena capitis lateralis (Watson) joined the one from the trigeminal

foramen at the termination of the groove (VG) for the latter vein at the dorsal

margin of the pterygo-paroccipital foramen where a distinct lamina of the

prootic protects the confluence (Figs 22A, 26).

It appears therefore that:

(a) the upper end of the third root of the vena capitis lateralis with its own root from

the posttemporal fossa is Cox’s vena capitis dorsalis;

(6) the lower end of the third root of the vena capitis lateralis which is not recognized

by Cox is present;

fossa to the pterygo-paroccipital foramen is the communicating vein of Cox, and

(d) the first root of the vena capitis lateralis which ran towards the trigeminal foramen

is the vena capitis lateralis of Cox.

The interpretations of Watson (1920), Cox (1959) and the present one for

Thrinaxodon are compared diagrammatically in Figure 23.

Parietal (Figs 6, 17A—20, 22B, 28, 29)

The relationships of the parietal with the frontal and postorbital have been

described above (see Epipterygoid, p. 367). In this region the parietal (PAR) is a

thin sheet of bone (Figs 17A—18C). Further back, behind the frontal, the parietal

is thicker in section and along its lower end, which almost reaches the upper

margin of the epipterygoid, runs the groove for the vena parietalis (Cox 1959)

(Fig. 18, GRVSC). Behind the parietal foramen, which is about 4 mm in length,

the two parietals are fused and appear in sections as a triradiate structure, the

dorsal limb of which forms the parietal crest (Figs 19, 20, 28, 29A, B). Still further

back the groove along the lower margin of the parietal disappears. The lower

margin of the parietal becomes rounded for a distance of about 6 mm and is

separated from the upper borders of the epipterygoid and prootic by an inter-

space of about | mm (Figs 19, 20, 29A) which may be the result of a dislodge-

ment of the bones, or may represent a growth feature which would have dis-

appeared later. Along the posterior dorsal part of the ‘sinus canal’ between the

parietal and prootic the considerably expanded lower end of the parietal and the

similarly expanded upper part of the prootic are again distinctly grooved (Fig.

29A, PAR, PRO). In this region the anterior end of the supraoccipital forms the

inner wall of the canal (Fig. 26) which now turns downwards and outwards.

Basioccipital and Exoccipital (Figs 1, 6, 16A, 19B-21, 25A, 27—29C)

Very little can be added to the previous descriptions by Olson (1944) and

Parrington (1946). Each exoccipital (EXO) meets the supraoccipital (SUOC)

in the sectioned skull in a sutura squamosa (Figs 6A, 28). An antero-ventral

extension of the exoccipital lies on the latero-dorsal surface of the posterior part

376 ANNALS OF THE SOUTH AFRICAN MUSEUM

PAR.

IPAR.

WPS.

PYC JU T.

SRV ly Ry OR

aa 1 \A SAN

i /i Ne Q

Nig

wth LLF. Xt

Fig. 29. Transverse sections through posterior part of braincase and occiput to

show relationships of constituent bones. Structures accommodating inner ear are

THE CRANIAL MORPHOLOGY OF THRINA XO DON LIORHINUS SEELEY EME!

of the basioccipital (Figs 6A, 27, 28, 29B, C, EXO, BO) where it forms the

inner wall of the jugular foramen. Two distinct foramina (Figs 1, 6A, F,,,)

penetrate the exoccipital. The lower one (LLF,,;;), which is the larger of the

two, leaves the braincase almost in the foramen magnum, and practically below

the jugular foramen. The smaller upper foramen lies along the posterior part of

the first foramen, but lateral to it, and may have given passage either to a second

hypoglossal branch, or to a vein (see Figs 27B, 28).

Fig. 30. A. Inner view of roof of snout of enlarged wax repro-

duction of serially sectioned skull to show series of ridges and

depressions probably for attachment of turbinals. B-—C.

Graphic reconstructions of medial views of postorbital and

tabular respectively. x 2,5.

Supraoccipital (Figs 1, 6A, 25A, 27-29)

Mention has been made of the dorso-lateral parts of the supraoccipital

(Figs 6A, 28, 29A, SUOC) which extend forward about 2,6 mm beyond the median

dorsal shelf on each side of the skull along the upper border of the prootic and

the lower border of the parietal. The supraoccipital is a trough-shaped bone

which lies tilted back at an angle of about 45 degrees (Figs 6A, 27, 28). The

base of the trough forms the roof of the cerebellar part of the braincase and the

walls of the trough, which become thicker backwards (Fig. 29A, B), take part in

the formation of the walls of the braincase. The anterior upper part of the roof

378 ANNALS OF THE SOUTH AFRICAN MUSEUM

is considerably expanded in vertical diameter (Figs 6A, 27B, 28). Above this

expanded anterior upper end of the supraoccipital, below and in front of the

interparietal is an open space about 3 mm in height and about 2 mm in length

(Figs 6A, 27B, 28, 29B, IPAR, SUOC). Its maximum anterior width is about

4,5 mm, but it becomes narrower further back. Such a space is also found in the

cynognathids investigated, as well as in Diademodon (Brink 1955a), but its func-

tion is uncertain. Brink suggests that in the live animal a cartilaginous extension

of the supraoccipital filled the space. The lower ends of the interparietal come

down laterally alongside the expanded anterior part of the supraoccipital shelf

(Figs 6A, 28, 29B, IPAR, SUOC) against which they are held by the parietal

flanges which completely cover them in lateral view (Figs 6B, 29B, IPAR, PAR).

The anterior extensions of the tabulars (Fig. 30C) lie below the lower ends of the

interparietal and medio-ventral to the lower ends of the parietals (Fig. 29B,

TAB). In lateral view they are exposed for a distance of about 2,2 mm before

being covered by the squamosal. This bone forms a roof (Figs 2, SQ, 26) which

probably protected the blood-vessels, emerging from the posttemporal fossa,

against constricting pressure of the temporal musculature. The anterior exposure

of the tabular in the posterior end of the groove for the ‘vein of the sinus canal’

is also found in the cynognathids and Diademodon.

About half-way down its sloping anterior border on each side, the supra-

occipital has a wide notch (Figs 6A, 27A, 28, SUOC). This notch is the posterior

medial margin of the excavation for the lateral lobe of the cerebellum which is

continued postero-laterally towards the posttemporal fossa between the pos-

terior part of the prootic and the supraoccipital. The posterior end of the anterior

vertical semicircular canal runs along the notch, but just behind its margin

(Fig. 28, AVC). The anterior upper end of the posterior vertical semicircular

canal runs in the outer surface of the lower end of the supraoccipital below

and behind the notch (Fig 28, PVC). A groove (Figs 6A, 24A, FSA) on the

inner upper surface of the posterior part of the prootic is separated from the

excavation for the lateral lobe of the cerebellum (LLCL) just below it by a

distinct ridge (Figs 6A, 24A, 27A). Medial to it the posterior part of the anterior

vertical semicircular canal runs along the margin of the supraoccipital notch

(Fig. 29A, B). This groove on the prootic apparently comes from the

subarcuate fossa proper and could be the equivalent of the vascular duct

described for Oligokyphus by Kihne (see p. 373).

The lateral ventral surfaces of the supraoccipital lie on the paroccipital

parts of the opisthotics (Fig. 29C, SUOC, POP) and its lower medial surfaces

lie on the exoccipitals (Figs 6A, 27, 28, 29C, SUOC, ORSCRFJ, OPO, EXO,

POP). These contacts are simple sutura squamosa. On the occiput the lower end

of the interparietal overlaps the medio-dorsal part of the supraoccipital and

the medial part of each tabular overlaps the lateral dorsal parts of the supra-

occipital (Figs 25A, 29C, IPAR, SUOC, TAB). On the occiput the tabular also

encloses (Fig. 25A, TAB, PTF) the posttemporal foramen as described by

Parrington (1946).

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 379

Opisthotic (Figs 1, 6, 24B, 25A, 27-29)

This bone forms the posterior part of the wall of the fenestra ovalis, and

the prootic and basioccipital form the anterior part of the wall. The opisthotic

separates the fenestra ovalis and jugular foramen of which it forms the anterior

and lateral walls (Figs 25A, 27, FO, FJ, OPO). Posteriorly it meets the exoccipital

(Figs 6, 25A, 27, 28, 29C, EXO) and postero-dorsally the supraoccipital (Figs 6,

25A, 27, 28, 29C, SUOC). The antero-dorsal part of the paroccipital process is

formed by the prootic, and the remaining major part of the process is formed by

the opisthotic (Figs 1, 25A, 29B, C, PRO, POP, OPO). The antero-dorsal face of

the paroccipital process is excavated probably for the passage of veins (see

Prootic above and Fig. 26) and the antero-ventral surface is also excavated to

form the middle ear cavity which accommodates the stapes (Figs 1, 25A). The

presence of a foramen of unknown significance featured by Estes (1961) and

facing laterally in the paroccipital part of the opisthotic about midway between

the posterior border of the fenestra ovalis and the antero-median border of

the jugular foramen is confirmed (Figs 1, 25A, FN,). Conceivably the internal

jugular vein left the skull through this foramen and not with nerves IX, X and

XI through the jugular foramen (foramen lacerum posterium).

Stapes (Figs 1, 19C, 25A, B, 26, 29A)

Both stapes are preserved and in both the stapedial foramen has a central

position. Parrington’s (1946) figure indicates that the stapedial foramen in his

specimen lay more towards the distal end of the stapes. He suggested that in the

young form the proximal surface of the head of the stapes was covered by

cartilage about 1 mm thick. That the serially ground skull is not that of a fully

grown animal is proved by the fact that much larger specimens have been

described by other authors. Broom (1938a) has described a specimen of 88 mm,

while Parrington’s (1936a) specimen I was 85 mm. The present author has

examined two skulls of more than 80 mm in length while the serially ground

skull was only 67 mm. (All were measured from the tip of the snout to the end

of the parietal crest with the possible exception of Broom’s (1938a) specimen

for which the maximum length may have been given.) It seems, therefore, that

the central position of the stapedial foramen must be attributed to the youth of

the specimen, and the consequent absence of an ossified cap on the head of the

stapes. The proximal end of the right stapes does not quite reach the fenestra

ovalis (Fig. 25A, FO), but if a cartilage cap were added it would fit into the

cup surrounding the fenestra. A small process at the distal end of the stapes

makes contact with the quadrate and this must represent a quadrate process

in its correct position. Another even smaller distal process closely approaches the

lower end of the groove in the squamosal which housed the external auditory

meatus (Figs 1, 25A, B,,>,3, QPS, ES). These processes are so delicate that their

preservation can only be expected in an undisturbed stapes. It is therefore con-

cluded that the right stapes is in the correct position and that its proximal end

must have been capped by cartilage in life. Conceivably, however, some rotation

380 ANNALS OF THE SOUTH AFRICAN MUSEUM

might have taken place soon after the death of the animal due to drying and

consequent shortening of muscles and ligaments which held the stapes in

position in the live animal.

The stapes on the left—although it appears to be in the natural position —

has actually been pushed into the cup around the fenestra ovalis by the medial

displacement of the quadrate and postdentary bones.

The anterior arm of the stapes (Figs 1, 25A, B, 3) is distinctly wider and

stronger than the posterior arm. A small dorsal process (Figs 1, 25A, By » 3

DPS), about 1 mm in length, curves upwards and slightly inwards from the

posterior corner of the distal end of the stapes, towards the antero-distal end of

the paroccipital process. A second process which the present author regards as

the quadrate process of the stapes (Figs 1, 25A, B, » 3 QPS) extends ventro-

laterally from the middle of the ventro-distal surface of the stapes to meet a

similar, but slightly larger process, here called the stapedial process of the

quadrate, which extends upwards from the postero-medial surface of the

quadrate immediately above the medial part of the condyle (Figs 1, 19C, 25A,

QPS, SPQ). A third very small process, here called the extrastapes (Fig 25A,

B, 2,3, ES), extends laterally from the distal surface of the stapes just in front

of the dorsal process. It approaches the lower inner end of the groove in the

squamosal for the external auditory meatus very closely. Ossification of the distal

region of the stapes and especially of the processes seems to be incomplete, but

there can be no doubt about their presence.

There is some difficulty in homologizing these processes with those described

by Parrington (1946, 1949). The process here described as the quadrate process

of the stapes may, if the stapes had rotated somewhat around its longitudinal

axis, be Parrington’s extrastapes. The dorsal process gives no difficulty, but the

small process at its base, described here as an extrastapes, may conceivably

be regarded as part of the dorsal process. The oblique flat face of the stapes at

its antero-distal corner above the quadrate process may be regarded as the point

of contact between the stapes and the quadrate.

Hopson’s description of the Thrinaxodon middle ear is essentially correct,

but it resembles the tritylodontid condition in structure, and probably function,

more extensively than he implies. As in tritylodontids like Bienotherium, the

stapes would be activated via the quadrate and articular in any jaw opening

movement, but the function of this mechanism is unknown although it is a

preadaptation to the mammalian condition.

Quadrate (Figs 1, 19B, C, 25A, 26, 29A, 31C, D)

Parrington (1946) described this bone very well and the present investigation

largely served to corroborate his description. A foramen which he found in the

quadrate ‘... median to the central thickening and a little way above the articula-

ting surface’ appears to be that of a small canal (Fig. 19C, QC) found in the

displaced left quadrate. The canal runs transversely through the central thicken-

ing of the quadrate. In the intact quadrate this canal emerges on its lateral face

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 381

just behind the medio-dorsally directed process at the ventral edge of the quadra-

tojugal. In its normal position (Fig. 26, Q) the quadrate lies in a pocket (Fig 26,

DQ) in the anterior face of the squamosal where it is held very loosely. The

quadrate has a posteriorly directed flange which fits into a deep notch in the

ventral border of the squamosal (Fig. 25A, QFL).

Olson (1944) and Parrington (1946) described a stapedial process of the

quadrate (Figs 1, 25A, SPQ) which makes contact with a process of the stapes

(see p. 379, Stapes). These authors have also shown that the posterior tips of the

quadrate ramus of the pterygoid (QRPT) and the lateral lamina of the prootic

(PROF) meet the medial surface of the quadrate (Figs 1, 19C, 26).

The quadrate is hollowed medially as in tritylodontids (Hopson 1966) and,

as in these forms, is not buttressed firmly against the squamosal, although it still

has no contact with the otic elements other than its medial contact with the

lateral lamina of the prootic (see Figs 19C, 29A).

Quadratojugal (Figs 1, 19B, C, 25A, 26, 29A, B)

The present investigation has served to elucidate some features of this bone

previously remarked on by Olson (1944) and Parrington (1946). The ventral end

of the quadratojugal projects below the ventral border of the squamosal (Fig.

25A, SQ, QJ) but not sufficiently far down to make contact with the post-

dentary bones of the lower jaw normally. Contact may, however, have been

established between the outer surfaces of the surangular and quadratojugal

when the jaw was opened very widely (Fig. 25A, SAN). The upper part of the

quadratojugal (Figs 1, 25A, QJFL) is expanded into a head, the dorsal surface

of which is grooved (Fig. 26, GHQJ; Fig. 29B, QJ). The foot of the quadrato-

jugal rests against the lateral surface of the quadrate at the base of the central

thickening immediately above the lateral part of the condyle (Figs 25A, 26).

The median surface of the vertical sheet of the quadratojugal (QJFL) and the

lateral surface of the quadrate flange (QFL) are concave (Fig. 26). An anteriorly

directed flange of the squamosal (Figs 1, 24C, 25A, 29A, SQS), which separates

the quadrate and quadratojugal flanges, lies in this space between the concave

surfaces of the two bones. A similar relationship is found in the cynognathids.

The ventral surface of the foot of the quadratojugal is convex and fits into a

concave quadrate surface (Figs 19B, 25A, 26, QJ, Q). The foot itself expands

lateral to the vertical flange of the quadratojugal and the ventro-lateral part of

the quadrate (Figs 1, 25A, 26). The lateral face of the quadratojugal flange is -

also concave, and another flange of the squamosal, expanded at its base, fits

into the upper end of this concavity and helps to hold the quadratojugal in

place (See e.g. Figs 1, 6B, 25A, 26).

Squamosal (Figs 1, 6B, 18A—20, 24C, 25A, 26, 29)

The present investigation added little to our knowledge of the squamosal.

The pocket for the dorsal flange of the quadrate in the anterior face of the

squamosal (Fig. 26, DQ) is only 1,6 mm deep. Another pocket (Fig. 26, DHQJ)

382 ANNALS OF THE SOUTH AFRICAN MUSEUM

lies immediately lateral to the first and housed the grooved head of the quadrato-

jugal (GHQJ). The anterior part of this pocket is confluent with that for the

quadrate, but the posterior deeper part is initially separated from it by a thin

wall of bone which expands posteriorly into the hind wall of the recess of the

quadrate. The total depth of the pocket is 3,4 mm (Fig. 29B, QJ). Its posterior

lower end is confluent with the deep notch for the quadratojugal flange in the

posterior lower border of the squamosal. In anterior view the quadratojugal

extends obliquely inwards and upwards into the pocket (see Fig. 26).

The tips of the quadrate ramus of the pterygoid (QRPT) and the lateral

lamina of the prootic (PROF) reach the quadrate just in front of the inner wall

of the pocket for the quadrate (Fig. 26). Crompton (1958: 191) has stated in

error that in Thrinaxodon the quadrate ramus of the pterygoid does not reach

the quadrate. In primitive cynodonts the quadrate rami of both the pterygoid

and the epipterygoid reach the quadrate. The advanced cynognathids no longer

retain the contact between the quadrate ramus of the pterygoid and the quadrate.

Cynognathus does retain a contact between the epipterygoid and the quadrate.

Cynidiognathus, which survived into Molteno Sandstone times (base of Storm-

berg Series overlying Beaufort Series), lacks this contact. Apparently the quadrate

ramus of the pterygoid was withdrawn in early cynodonts and the quadrate

ramus of the epipterygoid in late cynodonts.

Because the anterior surface of the paroccipital process is hollowed out,

the antero-distal part appears as a distinct forward extension. This extension

abuts against that part of the squamosal which forms the medial wall of the

pocket for the quadrate (Figs 26, 29A, POPA). Because the distal surface of the

paroccipital process does not make close contact with the squamosal (see e.g.

Figs 26, 29B, POP, SQ), it seems reasonable to assume that the distal surface

was covered by cartilage. Preservation of the distal end of the paroccipital

process in Cynidiognathus (and possibly other cynognathids) also gives the

impression that a cartilage cap may have been present. Such an arrangement

might have been necessitated by some cranial kinetic movement at this point.

A distinct flange of the squamosal overhangs the posttemporal fossa

(Fig. 2, PTF, SQ). This flange prevented constriction by the temporalis muscle

of the veins (see above, Prootic, p. 373) which passed through the posttemporal

fossa.

Orbitosphenoid

No trace of an orbitosphenoid was found in Thrinaxodon, but such an

element was found in Cynognathus and Cynidiognathus, and has been described

by Brink (1955a) in Diademodon.

Nasal septum and interorbital system

The presence during life of a nasal septum and an interorbital septum can

be inferred from the structure of the vomer, the pterygoid and the parasphenoid.

A groove runs along the median dorsal margins of the vomer and the para-

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 383

sphenoid (Fig. 6A, GR and all appropriate figures of sections). It seems probable

that a nasal septum rested on the vertical plate of the vomer, a cribiform

plate on the anterior part of the roof of the nasopharyngeal passage formed by

the horizontal plate of the vomer, the palatine and the pterygoid, and an inter-

orbital septum on the rostrum of the parasphenoid. No trace of an ossified

interorbital septum similar to that found in Cynognathus, Cynidiognathus, and

Diademodon was found. In these forms the median ossification is referred to as a

presphenoid.

The existence of a cartilaginous posterior wall to the nasal capsule was

suggested by Crompton (1958) for Diarthrognathus due to a difference in the

matrix filling of the region where this structure would be expected. He suggested

that such a difference in the matrix could be caused by finely suspended material

which permeated through the more coarse material anterior and posterior to the

space left by the disintegration of the cartilaginous wall. Possible evidence of a

similar process was found in the sectioned skull of Thrinaxodon. The matrix

in the nasal cavity itself is the same as that filling other cranial cavities and that

surrounding the skull. In the left side of the snout a matrix which resembles fine

quartz or calcite fills the space stretching from behind the posterior palatal

foramina up to the choanae. This type of matrix differs conspicuously from the

matrix found elsewhere in the skull.

Comparison with a model built of a sectioned skull of Erinaceus europeus

in the course of the present investigation shows that the position of this calcite-

like matrix plate between the posterior palatal foramina and the choanae agrees

exactly with the position occupied by a cribiform plate in a mammal. The prob-

lem is, however, that in mammals the secondary palate is more extensive

posteriorly, so that the choanae are further back than in cynodonts. If the cribi-

form plate was indeed situated here in cynodonts there would apparently have

been a gap between the posterior region of this plate and the interorbital septum

further back. As this appears unlikely, the calcite-like filling should probably

not be considered as an indication of a cribiform plate, or else its position is

not a precise indication of the position occupied by the structure, the presence

of which caused the calcite-like filling to develop. The presence of the calcite

on the left side only may be a result of differences in the rate of disintegration

of the possible cartilage.

Inner surface of the snout

A series of grooves and ridges on the inner surfaces of the nasals and

frontals (Fig. 30A) is identical to those described by Watson (1913a) in Nytho-

saurus and Diademodon. He considered these ridges and grooves, which closely

resemble those in mammals, to be associated with the attachment of ethmo-

turbinals. On the inner surfaces of the maxillaries, similar but less pronounced

ridges are found. These ridges may denote the presence of rudimentary maxillo-

turbinals. Similar grooves and ridges were also found in the cynognathids.

Because this feature is generally associated with homoiothermy it is interesting

384 ANNALS OF THE SOUTH AFRICAN MUSEUM

to find it even in early cynodonts, where the possibility of incipient homoio-

thermy therefore must be considered.

Internal ear (Figs 27—29C)

The description given by Olson (1944: 28-29) of the internal ear of Thrinaxo-

don liorhinus 1s confirmed by the present study. The sacculo-cochlear recess

(Figs 27A-29B, COE, CO, SR) is fairly big and ovoid in shape. The fenestra

ovalis is situated lateral to it. The recess lies on the same level as the floor of

the braincase. The major part of the outer border of the fenestra ovalis is formed

by the prootic, its posterior and part of its inner border by the opisthotic, and

the anterior inner part of its border by the basioccipital, which also forms part

of the medial border of the sacculo-cochlear recess. A thick bony wall of the

opisthotic (Fig. 28, ORSUSR) demarcates the sacculo-cochlear recess from the

utriculus. The separation of the sacculo-cochlear recess and the jugular foramen

by a ridge of the opisthotic (Fig. 28, ORSCRFJ) is incomplete. The ampullar

recesses (AAMP, HAMP, PAMP) and the courses of the semicircular canals

(Figs 28, 29, AVC, HC, PVC) are described accurately by Olson (1944) and

Estes (1961).

The channel identified by Estes as a fenestra rotunda, which joins the

fenestra ovalis and the jugular foramen, is very distinct in the Thrinaxodon skull

reconstructed from serial sections in the present investigation (Figs 25A, FNA,

27B). A reconstruction of the inner ear structures (Fig. 28) added little to the

descriptions by Olson and Estes. (See also under Supraoccipital, p. 378.)

(5) LOWER JAW (Figs 2, 31-33)

New information obtained by serial grinding of a very well-preserved skull

and lower jaw and reconstruction on an enlarged scale makes it necessary to

reconsider some features described and discussed by previous authors. These

new features pertain mainly to the angular, articular and dentary.

Angular (Figs 14, 15, 17A-19B, 20, 22B, 25A, 31C, D, 32)

The reflected lamina (ANRFL) of the angular is preserved on the left side

(Figs 17C, 31C, D, 32B). Here it is crescent-shaped, but broad as figured by

Olson (1944). On the right side it is broken off, but it is likely that a slender,

curved piece of bone found in the matrix on the medial side of the right ramus

is actually the lamina. Parrington (1946) described the structure as a broad sheet

of bone. He suggests that the anterior and lower parts, being more substantial

than the rest of the structure, tend to remain in specimens in which the delicate

sheet has been lost by weathering or preparation. Comparison with other

specimens in the National Museum, Bloemfontein, in which the reflected lamina

is preserved, indicates that the loss of the thin sheet, as Parrington suggests

usually happens, would give the structure a crescent shape.

Parrington’s (1955) investigations led him to conclude that the masseter in

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY

189

264 274 287 297 306

suenaneysaeeeneay

Bi 66

RLPC7 |UELPCg

64. 274. 287. 297. 306.

LPC7. > LPC8- COR SANS

cor SAN AN

yy Cor \—SAN

PA. PA )}-PA

PA. AN.

Sy AN PA L_AN

CMDN. “i REL

D. S ANREL.

131. 156. 189. 214. 231. 244

Fig. 31. A-B. Graphic reconstructions of anterior and medial views

respectively of retroarticular process of articular on right side of serially

sectioned skull. C. Graphic reconstruction of posterior part of lower jaw

in medial view to show form, position and relationships of comprising

elements. Important levels to show how dentary and postdentary bones

are joined together are indicated by the numbers of the sections of the

serially sectioned skull and lower jaw. D. Sections indicated in C. x 2.

385

386 ANNALS OF THE SOUTH AFRICAN MUSEUM

UELPCg.

i ,

at EY

pi Ome EY

DPS

Ke ae 4

5 Bi

\ a]: Fe

N Hi

.: i

f 4 Hi

it i

bh] S/d

f STS 7

Fig. 32. Graphic reconstructions of lower jaw. A. Dorsal view; B. Ventral view. The latter is

shown as a mirror image. x 1,5. -

gorgonopsids and in early cynodonts, like Thrinaxodon, was still inserted on to

the reflected lamina of the angular. He suggested that the transfer of the insertion

on to the dentary itself took place only in the advanced cynodonts with a large

dentary. He believed that only then could the final reduction of the reflected

lamina begin. Since the reflected lamina in Thrinaxodon is, however, already

very much smaller than in Gorgonopsia and Therocephalia, and even slightly

smaller than in procynosuchids, it appears probable that at least a large part of

the masseter muscle must already have shifted its insertion on to the angle of the

dentary. The very small reflected lamina figured by Broili & Schréder (1934a,

1935a) for Cynognathus, and the equally small one in Trirachodon specimens in

the National Museum and in Diademodon, give the impression of it being a dis-

appearing structure. It seems, therefore, that the reflected lamina had served

THE CRANIAL MORPHOLOGY OF THRINA XO DON LIORHINUS SEELEY 387

initially as insertion for the masseter muscle as Parrington (1955) concluded,

but that it was already becoming unnecessary in the Middle Beaufort cynodonts

where the masseter was already largely if not completely inserted on to the angle

of the dentary.

Crompton (19636) in his discussion of the origin of the mammalian lower

jaw and its musculature accepted Parrington’s suggestion that the masseter

transferred from the reflected lamina on to the angle of the dentary. He suggested

that there had been a superficial masseter inserting on the reflected lamina, and

a deep masseter inserting on the postdentary bones. Migration of these muscles

on to the dentary would have caused the reduction of the elements on to which

they inserted.

Barghusen (1968) rejects these interpretations. On the basis of modern

reptiles he believes that three muscles may originally have attached to the

angular keel and massive reflected lamina in sphenacodonts and early therapsids.

Of these three muscles he considers the branchiomandibularis as the most likely

to have retained its attachment to the reflected lamina in cynodonts after the

possible earlier pterygoideus musculature had been lost and the intermandibu-

laris had shifted entirely on to the enlarged dentary, causing the reduction of

the reflected lamina. He also concluded that the cynodonts indeed developed a

masseter-like muscle, but ina manner different from that proposed by Parrington

or Crompton. He believes that it differentiated from the external adductor and

achieved insertion on the lateral surface of the dentary in primitive cynodonts.

This insertion then migrated downwards to the postero-ventral edge of the

dentary in later cynodonts. This development is uniquely cynodont and closely

approaches the adductor jaw musculature of mammals. This is one of the reasons

why cynodonts appear better fitted than any other group as the ancestors for

all mammals.

Articular (Figs 19, 20, 31, 32)

The articular (AR) is held closely between the posterior lower half of the

surangular (SAN) and the posterior part of the prearticular (PA) medially

(Figs 31C, Do74 987, 297, 32). It is a short solid bone with a pronounced bulge

(Fig. 32A) towards the prearticular. Its median-dorsal surface is concave (Fig.

31D.7), a feature which is more pronounced along the central part of the bone

than along its posterior part.

This concave median-dorsal surface results in a latero-dorsal ridge (Fig.

31B, Dysg7), which is made more pronounced because the vertical diameter of

the bone is larger along the lateral margin of the bone than along the medial

bulge. This bulge becomes progressively less distinct towards the posterior end

of the bone, where it ends as a ridge on the upper medial surface of the retro-

articular process (RAPA) (Fig. 31B, Dy97 396), Another ridge only becomes

perceptible behind the central part of the bone. It runs backwards from the

ventro-lateral surface (where it originates as a result of the medially directed

bulge) to a ventro-median position, becoming progressively more distinct on

388 ANNALS OF THE SOUTH AFRICAN MUSEUM

its way. Finally it terminates in the retroarticular process which is deflected

downwards and forwards (Figs 31A, Dogz 997 306, 32B).

The presence of a well-developed retroarticular process in Thrinaxodon

has apparently not been described previously. Such a process is preserved only

on the right articular of the serially ground skull (Figs 31A, B, C, Do. 32B,

RAPA) and is missing in all the other Thrinaxodon skulls in the National

Museum. Apparently the retroarticular process in Thrinaxodon is particularly

apt to break off either prior to fossilization or later during exposure and

weathering.

Parrington (1955) discusses the mechanics of retroarticular processes.

He examines the deflected articular process found in synapsids generally, and

concludes that this deflected process is a true retroarticular process which gave

attachment to a depressor mandibuli muscle in all the synapsids. This conclusion

contradicts that of Watson (1948, 1951) who believes that in late synapsids the

true retroarticular process had for the most part disappeared, and with it the

depressor mandibuli muscle. According to Watson the posterior pterygoid

muscle attached to the structure, which Parrington calls a retroarticular process

in the late synapsids, and sublingual muscles (or muscles comparable with those

of mammals) opened the jaw. Parrington shows that the functional implications

of this determination raises grounds for doubting Watson’s interpretations. An

interpretation of the disputed structure and its modifications as a retroarticular

process allows for a straightforward explanation in terms of jaw-mechanics.

Dentary and splenial (Figs 2-4, 8, 9, 11, 14, 15, 17A-18B, 22B, 31C, D, 32)

Each lower jaw ramus is formed mainly by the dentary which is joined by

an interdigitated suture only at its extreme anterior end to the dentary of the

opposite ramus (Fig. 32, SY). The posterior part of the lower jaw symphysis is

formed by an interdigitated suture between the two opposite splenials (Fig.

32, S). The anterior part of the splenial which takes part in the symphysis, is

thick and solid (Fig. 8A, S) compared with the rest of the bone which has a thin

upper and somewhat thicker lower part (Figs 9, 11, 14, S). The splenial lies

against the lower half of the medial surface of the dentary where the expanded

lower end of the splenial fits into an excavation along the lower part of the

dentary.

Due to the loss of the incisor bearing part of the mandible, the mandibular

canal (Figs 8-10, CMDN) for the ramus mandibularis of the trigeminal nerve

could not be traced further forward than the canine alveolus into the base of

which it enters. The major part of the canal anterior to the seventh postcanine

lies entirely in the dentary. It runs just below the bases of the postcanine alveoli,

with each of which it retains contact. Behind the seventh alveolus the medial wall

of the canal is formed by the splenial. The anterior end of the prearticular, which

is inserted between the medial surface of the dentary and the upper part of the

splenial, extends forward to this level where the mandibular canal enters the

dentary (Figs 11B, 31C, D,3;, 32, PA). The tip of the prearticular lies just above

THE CRANIAL MORPHOLOGY OF THRINAXO DON LIORHINUS SEELEY 389

the mandibular canal. This bone, which is very thin, is expanded dorso-ventrally

for some distance behind the level of the eighth lower postcanine (Fig. 31C, PA).

Its lower end reaches down below the upper medial dentary margin of the

mandibular canal. Thus it forms the upper part of the medial wall of the pos-

terior part of this canal into which the anterior end of the angular penetrates

as far as the eighth postcanine (Fig. 31C, D,;,, AN). Above the mandibular canal

behind the seventh postcanine, the medial surface of the dentary is slightly

concave, apparently to aid in the attachment to the dentary of the prearticular

and splenial, and further back the coronoid (Figs 14-15B, D, S, PA, COR).

The channel in the dentary behind the seventh postcanine becomes progressively

higher. The medial part of its upper wall is formed by the lower surface of the

tooth-bearing ridge (Figs 11B, 14A, B). Where this structure ends well behind

the anterior border of the coronoid process of the dentary, a medio-ventrally

directed ledge forms the upper wall of the channel for approximately a further

5 mm. Behind this ledge there is no further indication of the mandibular

canal.

The anterior end of the angular which penetrates into the mandibular canal

divides it into an upper and a lower half. Along the dorsal surface of the angular

runs a groove of which the lateral wall is considerably higher than the medial

one. This groove extends as far back as the angle of the dentary (Figs 14, 15, D,

AN). Still further back the reflected lamina (ANRFL) is attached to the base

(Fig. 17B, C) of the angular and immediately behind this level the angular is

considerably expanded in a dorsal direction (Figs 31C, Do4,, AN). This part of

the bone is tilted laterally, with its dorsal margin directly below the lower margin

of the coronoid process of the dentary. This part of the angular has a peculiarly

contoured shape to maintain a close relationship between itself, the prearticular

and the surangular. This latter bone lies against the inner surface of the upper

half of the angular which is curved in such a way as to form a ledge for the

support of the surangular, the posterior half of which is expanded in a ventral

direction (Figs 31C, Dogs, 274, 32A, AN, SAN). The lower margin of the posterior

half of the angular is curved medio-dorsally and fits into a groove along the

ventral surface of the posterior part of the prearticular (Figs 31 Doga, 274, 287, 32B,

PA) thus forming a schindylesis or wedge-and-groove suture. Anterior to the

reflected lamina of the angular, the prearticular lies at the medio-dorsal margin

of the angular. Because the lateral surface of the prearticular is concave, a chan-

nel is formed between the concave dorsal surface of the angular and the concave

lateral surface of the prearticular (Figs 31D 55 71, 32A, AN, PA).

The articular inserts into the posterior part of this channel (Figs 31Do71,

287, 297, 32A, AR). Its anterior dorsal surface is concave (grooved) and covered

laterally by a peculiar ledge of the surangular (Figs 31C, Do7a 987, 32A, SAN).

A possible course for the ramus mandibularis of the trigeminal can now be

suggested. It probably ran backwards in the mandibular canal; then along the

dorsal surface of the angular where the latter penetrates into the canal; then

along the channel formed by the dorsal and further back the medio-dorsal

390 ANNALS OF THE SOUTH AFRICAN MUSEUM

surface of the angular and the lateral surface of the prearticular; then along the

dorsal surface of the anterior part of the articular which inserts into the angular-

prearticular channel. From the articular, but in front of the quadrate, the ramus

mandibularis probably passed sharply medianly and dorsally in front of the

stapes. At its base the lateral lamina of the prootic is notched on its anterior

margin. Passing through this notch the ramus mandibularis then probably

entered into the cranial cavity via the large foramen between the prootic and

epipterygoid.

Coronoid (Figs 11B, 14A—-15B, 31C, D, 32A)

The tip of the coronoid lies just behind the sixth postcanine. The bone

extends along the medio-dorsal surface of the dentary above the splenial, pre-

articular and angular. In this region the upper part of the coronoid lies above the

dorsal margin of that part of the dentary which forms the medial wall of the

eighth postcanine alveolus. The ninth postcanine alveolus, which is partly

formed, has a crown of an unerupted tooth of much smaller size than the eighth

in it. The dentary has hardly any part in the formation of the medial wall of the

ninth alveolus. The coronoid covers the partly formed ninth alveolus medially

(Figs 14B, 31C, 32A, COR, UELPC,). Behind this ninth alveolus the coronoid

increases considerably in height (Figs 31C, 32A, COR). Its upper part is thicker

than its lower edge which covers the upper edge of the prearticular (Figs 15A, B,

31C, COR, PA). The medial surface of the coronoid process of the dentary

behind and above the postcanine alveoli is concave to receive the upper part of

the coronoid (Figs 15A, B, 31D,.5, D, COR). Just behind the tip of the surangular

which is inserted between the coronoid and the dentary, the coronoid bifurcates

into an upper part which ends abruptly and a lower part which extends back-

wards for about 3,5 mm along the upper part of the medial surface of the pre-

articular (Figs 31C, 32A, COR, PA).

Surangular (Figs 17A-19B, 20, 22B, 25A, 31C, D, 32)

The surangular has a long anterior portion. This part of the bone is very

thin and lies in a shallow depression on the medial surface of the coronoid

process of the dentary. On the lateral surface of this coronoid process is a ridge,

the position of which (Fig. 31Do44, 931, 244, D, SAN) corresponds with the medial

depression for the surangular. This ridge, which runs obliquely upwards towards

the posterior margin of the dentary, forms the lower margin of the insertion

area of the masseter muscle.

Towards and beyond the posterior margin of the coronoid process of the

dentary the long anterior and the posterior parts of the surangular become pro-

gressively thicker. Well behind the coronoid process and above the reflected

lamina of the angular, the surangular bears an antero-ventrally directed process,

2 mm long, below the long anterior part of the bone. The base of this process is

that part of the bone which is expanded in a ventral direction and lies medial

to that part of the angular which is expanded in a dorsal direction (Figs 31C,

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 39]

32A, SAN, AN). The relationships of the angular and surangular have been

described above (see p. 389).

Some reference has been made to the relationship between the surangular

and the articular (see p. 387). The lower edge of the surangular lies against the

lateral surface of the articular. Less than 2 mm anterior to, but above the tip of

the articular, a ridge begins on the medial surface of the surangular (Figs 31C,

32A, AR, SAN). Further back this ridge progressively expands into a ledge

running above the lateral part of the articular. The upper half of the surangular,

which has a concave medial face and a thickened dorsal edge, ends on the ledge

(Figs 31C, Dogs, 274, 287, 32A, SAN). The ledge itself ends just behind the bulky

anterior part of the articular, and the thin lower edge of the bone ends along the

flattened neck about half-way between the anterior part of the articular and the

retroarticular process (Figs 31C, Dyg,, 322A, AR, SAN, RAPA).

Prearticular (Figs 11B, 14A—15C, 17A-19B, 20, 22B, 25A, 31C, D, 32)

The relationships of the prearticular with the dentary and splenial (see

p. 388) and with the angular (see p. 389) have been described above. It can also

be noted that the posterior part of the prearticular, of which the ventral margin

is grooved for insertion of the medio-dorsally curved ventral margin of the

angular, is increasingly tilted medially further back where it is a broad thin plate.

This structure is correlated with the rapid progressive increase in diameter of

the anterior part of the articular of which the medial surface is covered by the

prearticular which closely follows its contours (Figs 31Dgyg7, 257, 32A, AR, PA).

Only between the postero-dorsal part of the medial face of the articular, and the

postero-dorsal part of the lateral face of the prearticular is there an interspaec

which may be an indication of the presence of the tensor tympani muscle (Fig.

31D,97). In mammals this muscle is typically inserted on to the medial surface

of the malleus, which is considered homologous with the articular and runs

forward to its origin on the base of the skull.

A detailed description of the dentition of Thrinaxodon liorhinus is deemed

unnecessary because of the work done by Crompton and co-workers in recent

years. However, since they limited their studies to the postcanines, a few points

of interest about the incisors and canines revealed by the study on which this

paper is based may be mentioned.

Specimens actually investigated by the present author or described in the

literature reveal evidence of at least four replacements of the incisors between

the 35 mm and the 88 mm skull-length stages. Crompton (1955c) has pointed

out that in the cynodonts investigated by him replacement of incisors and canines

is found much more frequently than replacement of postcanines. This pheno-

menon is due, he believes, to the fact that incisors and canines form an arc

whereas postcanines form a linear series. During growth those parts of the

392 ANNALS OF THE SOUTH AFRICAN MUSEUM

foomat

Fig. 33. A. Graphic reconstruction to show relative

positions of replaced, functional and replacing upper

canines. B—C. Graphic reconstructions of lower and

upper dentitions. Only outlines of teeth crowns are

shown and replacing postcanines are shown in their

true position with respect to the functional teeth.

Lost incisors are indicated by dotted outlines only.

Alveoli of lost replacing teeth are shown in black.

Ae XxX? 2.5: BO << ali2a)

mandible and upper jaw containing incisors and canines increase greatly in size.

Frequent replacement of the functional teeth by larger ones would, therefore,

ensure that functional teeth occupied the entire region.

The current investigation of Thrinaxodon and the later cynodonts shows

that the type of replacement of the canines in these forms is identical with that

found by Kermack (1956) in the therocephalians and gorgonopsians which he

investigated. Although the crowns of the canines are apparently shed, the roots

are retained (Figs 3, 33A, (R)RC') and are gradually resorbed, thus apparently

aiding the calcium (and phosphate?) metabolism of the animal. There is some

evidence that the alternation of consecutive generations of functional canines

between the anterior and posterior halves of the alveolus, observed by Kermack,

can also take place in Thrinaxodon. One example of an upper replacing canine

THE CRANIAL MORPHOLOGY OF THRINAXO DON LIORHINUS SEELEY 393

erupting posterolingual to the functional one was found. In all other specimens

investigated the replacing canine was anterolingual to the functional one. Evi-

dence of the bone plug which closes the unoccupied part of the alveolus until

the new canine erupts, was found in the lower canine alveoli of the sectioned

skull (Fig. 3A, BP). Consideration of all available evidence suggests at least six

and probably seven or more replacements of the functional canines.

SUMMARY

1. A skull of Thrinaxodon liorhinus has been serially sectioned at intervals of

200u, and an enlarged wax model built. New details on the structure of the

component bones of the skull and their relationships are described and figured.

The processes of the premaxillary and their contacts with the maxillary are

described and figured. The canals in the maxillary are described and functional

and replacement teeth are indicated. The canals in the lacrimal and the relation-

ships of this bone with the maxillary and palatine to form the maxillary antrum

are indicated. Grooves and ridges on the inner surface of the roof of the snout,

possibly for the attachment of turbinals, are figured.

The vomer with a groove along its dorsal margin is described and it is

shown that the double anterior end suggests a paired origin of the bone. The

relationships of the premaxillary with the lower end of the vomer are indicated

as well as a possible explanation for the thickened lower margin of the anterior

vertical part of the bone and the medial dorsal margins of the maxillaries. An

explanation is also offered for the ridges on the ventral surface of the palatine

and pterygoid. The exact relationship of the posterior parts of the pterygoids and

the parasphenoid rostrum is indicated. It is shown that interdigitated sutures

preclude the possibility of kinetic movement in this region. It is suggested that

the groove dorsally along the vomer, and on the dorsal surface of the para-

sphenoid rostrum, may indicate the presence of a cartilaginous nasal septum and

an interorbital septum. Evidence in the matrix of a possible cartilaginous cribi-

form plate at the posterior end of the snout is described. The relationship of the

anterior ventral part of the epipterygoid and the pterygoid is indicated as well as

that of the epipterygoid and prootic. The walls of the braincase are described in

detail, externally and internally. The interpretations of Watson and Cox of the

venous system on the walls of the braincase in cynodonts and anomodonts are

compared with conditions in Thrinaxodon and an interpretation of the condition

in this form is suggested.

The structures of the prootic and opisthotic to accommodate the internal

ear structures are described and figured. The presence of a possible fenestra

rotunda is confirmed. The exact way in which the abducent foramen is formed is

described. The possible courses of the maxillary, mandibular and deep ophthal-

mic rami of the trigeminal, the abducent and the palatine ramus of the facial

nerves are described. A possible Eustachian foramen is indicated as well as a

possible separate foramen for the internal jugular vein. The basicranial region

is described and it is shown that the sella turcica is very shallow and that there

394 ANNALS OF THE SOUTH AFRICAN MUSEUM

is hardly any dorsum sellae. The stapes is described in detail and it is shown

that it has three processes on its distal end; these are a dorsal process towards

the paroccipital, a quadrate process, and, as preserved, a minute process

towards the base of the external auditory meatus. The relationships of the

quadrate and quadratojugal with each other as described by Parrington (1946)

are confirmed and their relationships with the squamosal are described.

Parrington’s (1946) description of a canal penetrating the quadrate is confirmed.

A stapedial process on the quadrate is described. The elements of the lower jaw

are described in detail and their relationships with each other are figured in

different views. A retroarticular process is described for the articular.

2. It is shown that the incisors are replaced at least four times and that during

replacement the crown of the functional canine is shed, but that the root is

gradually resorbed. Only one example of an upper replacing canine erupting

postero-lingual to the functional one was found. In all other specimens investi-

gated the erupting replacing canine was in the antero-lingual position in relation

to the functional canine. More definite evidence of the alternation of the func-

tional canine between the anterior and posterior halves of the alveolus seems

to be needed.

ACKNOWLEDGEMENTS

I wish to thank the Board of Trustees of the National Museum in Bloem-

fontein for the facilities and the Thrinaxodon skull provided for this study. My

sincere thanks are also extended to Dr A. W. Crompton, at present Director of

the Harvard Museum of Comparative Zoology, who originally suggested the

study and took a keen interest in it while he was Director of the South African

Museum in Cape Town. Both he and Professor M. E. Malan spent many hours

of their valuable time on the thesis, part of which was made up by this study in

its original form. Finally I wish to express my indebtedness to Dr T. H. Barry,

Director, and the Board of Trustees of the South African Museum in Cape Town

who made it possible for this study to be published.

REFERENCES

BARGHUSEN, H. R. 1968. The lower jaw of cynodonts (Reptilia, Therapsida) and the evolu-

tionary origin of mammal-like adductor jaw musculature. Postilla 116: 1-49.

BRINK, A. S. 1954. Thrinaxodon and some other Lystrosaurus zone cynodonts in the collection

of the National Museum, Bloemfontein. Navors. nas. Mus., Bloemfontein 1: 115-125.

BRINK, A. S. 1955a. A study on the skeleton of Diademodon. Palaeont. afr. 3: 3-39.

BRINK, A. S. 1955b. Note on a very tiny specimen of Thrinaxodon liorhinus. Palaeont. afr.

3: 73-76.

BRINK, A. S. 1957. Speculation on some advanced mammalian characteristics in the higher

mammal-like reptiles. Palaeont. afr. 4: 77-96.

BRINK, A. S. 1961a. A new type of primitive cynodont. Palaeont. afr. 7: 119-154.

BRINK, A. S. 19615. On some small therocephalians. Palaeont. afr. 7: 155-182.

Brink, A. S. 1965. On two new specimens of Lystrosaurus zone cynodonts. Palaeont. afr. 9:

107-122.

Brorut, F. & ScHROpER, J. 1934a. Zur Osteologie des Kopfes von Cynognathus. Sber. bayer.

Akad. Wiss. 1934: 95-118.

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 395

BroiLt, F. & ScurOpeER, J. 19346. Uber den Cynodontier Tribolodon frerensis Seeley. Sber.

bayer. Akad. Wiss. 1934: 163-177.

Bro, F. & SCHRGODER, J. 1935a. Uber den Schidel von Gomphognathus Seeley. Sber. bayer.

Akad. Wiss. 1935: 115-182.

Brom, F. & ScHROpER, J. 19356. Uber den Schadel von Cynidiognathus Haughton. Sber.

bayer. Akad. Wiss. 1935: 199-222.

Broom, R. 1903. On the classification of the theriodonts and their allies. Rep. S. Afr. Ass.

Advmt Sci. 1: 286-294.

Broom, R. 1910. Observations on some new specimens of South African reptiles preserved in

the British Museum. Trans. R. Soc. S. Afr. 2: 19-25.

Broom, R. 1911. On the structure of the skull in cynodont reptiles. Proc. zool. Soc. Lond. 81:

893-925.

Broom, R. 1912. On some new fossil reptiles from the Permian and Triassic beds of South

Africa. Proc. zool. Soc. Lond. 82: 859-876.

Broom, R. 1913a. On evidence of a mammal-like dental succession in the cynodont reptiles.

Bull. Am. Mus. nat. Hist. 32: 465-468.

Broom, R. 19135. On some new carnivorous therapsids. Bull. Am. Mus. nat. Hist. 32: 557-561.

Broom, R. 1925. On some carnivorous therapsids. Rec. Albany Mus. 3: 309-326.

Broom, R. 1931. Notices on some new genera and species of Karroo fossil reptiles. Rec.

Albany Mus. 4: 161-166.

Broom, R. 1932. The mammal-like reptiles of South Africa and the origin of mammals. London:

Witherby.

Broom, R. 1938a. On the structure of the skull of the cynodont, Thrinaxodon liorhinus Seeley.

Ann. Transy. Mus. 19: 263-269.

Broom, R. 1938b. The origin of the cynodonts. Ann. Transv. Mus. 19: 279-288.

Cox, C. B. 1959. On the anatomy of a new dicynodont genus with evidence on the position of

the tympanum. Proc. zool. Soc. Lond. 132: 321-367.

Crompton, A. W. 1955a. Techniques for the study of Permo-Triassic fossils of South Africa.

Bull. S. Afr. Mus. Ass. (Samab) 6: 57-60.

Crompton, A. W. 1955b. A revision of the Scaloposauridae with special reference to kinetism

in this family. Navors. nas. Mus., Bloemfontein 1: 149-183.

Crompton, A. W. 1955c. On some Triassic cynodonts from Tanganyika. Proc. zool. Soc. Lond.

125: 617-669.

Crompton, A. W. 1958. The cranial morphology of a new genus and species of ictidosaurian.

Proc. zool. Soc. Lond. 130: 183-216.

CroMPTON, A. W. 1963a. Tooth replacement in the cynodont Thrinaxodon liorhinus Seeley.

Ann. S. Afr. Mus. 46: 479-521.

Crompton, A. W. 1963b. On the lower jaw of Diarthrognathus broomi and the origin of the

mammalian lower jaw. Proc. zool. Soc. Lond. 140: 697-753.

Crompton, A. W. & JENKINS, F. A. 1968. Molar occlusion in late Triassic mammals. Biol. Rev.

43: 427-458.

DE Beer, G. R. 1937. The development of the vertebrate skull. Oxford: Clarendon Press.

Estes, R. 1961. Cranial anatomy of the cynodont reptile Thrinaxodon liorhinus. Bull. Mus.

comp. Zool. Harv. 125: 165-180.

GreGory, W. K. 1920. A review of the lachrymal bone of vertebrates with special reference

to that of mammals. Bull. Am. Mus. nat. Hist. 42: 99-263.

HAUGHTON, S. H. 1920. On the genus Ictidopsis. Ann. Durban Mus. 2: 243-246.

HAUGHTON, S. H. 1924. On Cynodontia from the Middle Beaufort beds of Harrismith, Orange

Free State. Ann. Transv. Mus. 11: 74-92.

Hopson, J. A. 1966. The origin of the mammalian middle ear. Am. Zool. 6: 437-450.

Hopson, J. A. 1967. Mammal-like reptiles and the origin of mammals. Discovery, Yale 2:

25-33.

Hopson, J. A. & Crompton, A. W. 1969. Origin of mammals. Jn; DoBZHANSKY, T., HECHT,

M. K. & STEERE, W. C., eds. Evolutionary biology 3: 15-71. New York: Appleton-Century-

Crofts.

JENKINS, F. A. 1971. The postcranial skeleton of African cynodonts. Bull. Peabody Mus. nat.

Hist. 36: 1-216.

KERMACK, K. A. 1956. Tooth replacement in mammal-like reptiles of the suborders Gorgonop-

sia and Therocephalia. Phil. Trans. R. Soc. (B) 240: 95-133.

396 ANNALS OF THE SOUTH AFRICAN MUSEUM

KUHNE, W. G. 1956. The Liassic therapsid Oligokyphus. London: British Museum (Natural

History).

LYDEKKER, R. 1890. Catalogue of the fossil Reptilia and Amphibia in the British Museum

(Natural History). 4. London: Taylor & Francis.

Oxson, E. C. 1944. Origin of mammals based upon cranial morphology of the therapsid

' suborders. Spec. Pap. geol. Soc. Am. 55: 1-136.

OSBORN, J. W. & CROMPTON, A. W. 1973. The evolution of mammalian from reptilian denti-

tions. Breviora 399: 1-18.

Ow EN, R. 1861. Palaeontology. 2nd ed. Edinburgh: Black.

OwEN, R. 1876. Descriptive and illustrated catalogue of fossil Reptilia of South Africa in the

collection of the British Museum. London: Taylor & Francis.

OwEN, R. 1887. On the skull and dentition of a Triassic saurian, (Galesaurus planiceps Owen).

Q. JI geol. Soc. Lond. 43: 1-6.

PARRINGTON, F. R. 1933. On the cynodont reptile Thrinaxodon liorhinus Seeley. Ann. Mag.

nat. Hist. (10) 11: 16-24.

PARRINGTON, F. R. 1935. A note on the parasphenoid of the cynodont Thrinaxodon liorhinus

Seeley. Ann. Mag. nat. Hist. (10) 16: 399-401.

PARRINGTON, F. R. 1936a. On the tooth-replacement in theriodont reptiles. Phil. Trans. R. Soc.

(B) 226: 121-142.

PARRINGTON, F. R. 19366. Further notes on tooth-replacement. Ann. Mag. nat. Hist. (10) 18:

109-116.

PARRINGTON, F. R. 1939. On the digital formula of theriodont reptiles. Ann. Mag. nat. Hist.

(11) 3: 209-214.

PARRINGTON, F. R. 1946. On the cranial anatomy of cynodonts. Proc. zool. Soc. Lond. 116:

181-197.

PARRINGTON, F. R. 1947. On the collection of Rhaetic mammalian teeth. Proc. zool. Soc. Lond.

116: 707-728.

PARRINGTON, F. R. 1949. Remarks on a theory of the evolution of the tetrapod middle ear.

J. Lar. Otol. 63: 580-595.

PARRINGTON, F. R. 1955. On the cranial anatomy of some gorgonopsids and the synapsid

middle ear. Proc. zool. Soc. Lond. 125: 1-40.

PARRINGTON, F. R. & WESTOLL, T. S. 1940. On the evolution of the mammalian palate. Phil.

Trans. R. Soc. (B) 230: 305-355.

Romer, A. S. 1956. Osteology of the reptiles. Chicago: University of Chicago Press.

SEELEY, H. G. 1889. On the anomodont Reptilia and their allies. Phil. Trans. R. Soc. (B) 180:

215-296.

SEELEY, H. G. 1895a. On the Therosuchia. Phil. Trans. R. Soc. (B) 183: 311-370.

SEELEY, H. G. 18955. On the skeleton in new Cynodontia from the Karroo rocks. Phil. Trans.

R. Soc. (B) 186: 59-148.

SEELEY, H. G. 1908. On the dentition of the palate in the South African fossil reptile genus

Cynognathus. Geol. Mag. (5) 5: 486-491.

SoOLLAS, I. B. J. & SoLLAs, W. J. 1914. A study of the skull of a Dicynodon by means of serial

sections. Phil. Trans. R. Soc. (B) 204: 201-225.

VAN HEERDEN, J. 1972. Intraspecific variations and growth changes in the cynodont reptile

Thrinaxodon liorhinus. Navors. nas. Mus., Bloemfontein 2: 307-347.

VAN Hoepen, E. C. N. 1916. Preliminary notice of new reptiles of the Karroo formation.

Ann. Transyv. Mus. 5 (3, suppl. 2): 1-2.

VAN VALEN, L. 1960. Therapsids as mammals. Evolution, Lancaster, Pa. 14: 304-313.

Watson, D. M.S. 1911. The skull of Diademodon with notes on those of some other cynodonts.

Ann. Mag. nat. Hist. (8) 8: 293-330.

Watson, D. M. S. 1913a. Further notes on the skull, brain and organs of special sense of

Diademodon. Ann. Mag. nat. Hist. (8) 12: 217-228.

Watson, D. M. S. 19135. On a new cynodont from the Stormberg. Geol. Mag. (5) 10: 145-148.

Watson, D. M. S. 1920. On the Cynodontia. Ann. Mag. nat. Hist. (9) 6: 506-524.

Watson, D. M. S. 1931. On the skeleton of a bauriamorph reptile. Proc. zool. Soc. Lond.

1931: 1163-1205.

WATSON, D. M. S. 1948. Dicynodon and its allies. Proc. zool. Soc. Lond. 118: 823-877.

WATSON, D. M. S. 1951. Palaeontology and modern biology. New Haven: Yale University Press.

BQREPT

CROP,

EAM

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 397

KEY TQ, LETTERING

anterior ampulla

antero-dorsal process of the prootic

angular

reflected lamina of the angular

articular

attachment of the rectus capitis anticus muscle

anterior vertical semicircular canal

premaxillary part of the anterior wall of the excavation for the lower canine

basicranial process

basioccipital

bone plug

basipterygoid process of the basisphenoid

basal plate of the epipterygoid resting on the basipterygoid process

broken quadrate ramus of the epipterygoid

continuation on to the basisphenoid of the ridge formed by the outer prong of

the pterygoid.

basisphenoid

broken symphysis

crus communis

carotid canal

cavum epitericum

choanae

canal for the mandibular nerve

canal for the maxillary nerve

cochlea

cochlear extension

coronoid

damaged coronoid process of the dentary

canal for the palatine ramus of the facial nerve

probable course of the maxillary ramus of the trigeminal nerve

probable course of the ramus ophthalmicus profundus of the trigeminal nerve

probable course of the palatine ramus of the facial nerve

vena communicans between vena capitis dorsalis and vena capitis lateralis

dentary

displaced fragment

displaced fragment of the transverse flange of the pterygoid

depression for a gland

depression on the squamosal for the head of the quadratojugal

duct for the lacrimal branch of the maxillary nerve

dorsal

dorsal process of the stapes

depression for the quadrate on the squamosal

displaced second replacing upper postcanine

displaced tooth

double termination of the canal for the maxillary branch of the trigeminal nerve

double anterior end of the vomer

external auditory meatus

entrance of the duct for the lacrimal branch of the maxillary ramus of the

trigeminal nerve

entrance of the independent maxillary canal

excavation for the canine of the lower jaw

entrance into lacrimal duct

epipterygoid

ectopterygoid

erupting replacing canine

erupting replacing second upper postcanine

extrastapes

FIMC

ICF

IDIOPPT

IMC

INPPX

IPAR

IPPT

ANNALS OF THE SOUTH AFRICAN MUSEUM

erupting upper postcanine

exoccipital

functional canine

foramina of double termination of canal for the maxillary ramus of the

trigeminal nerve

foramen in the ectopterygoid

foramen for trigeminal nerve

foramen for abducent nerve

foramen for facial nerve

foramen for hypoglossal nerve

frontal

displaced main fragment of the developing ninth lower postcanine

foramen of the independent maxillary canal

foramen incisivum

foramen jugulare

foramen lacerum anterium

foramen for the lacrimal branch of the maxillary ramus of the trigeminal nerve

foramen magnum

foramen for Eustachian tube (7)

foramen for internal jugular vein (7)

fenestra rotunda

foramen of canal to Organ of Jacobson and nasal gland

fenestra ovalis

fragment of the paroccipital process

foramen pterygoideum

fracture

fossa subarcuata

foramen for the vena capitis dorsalis in the sinus canal

groove along the dorsal margin of the vertical plate of the vomer

grooved head of the quadratojugal

groove

groove for the palatine ramus of the facial nerve

groove for the vein of the sinus canal

ampulla for the horizontal semicircular canal

internal carotid foramen.

excavation on pterygoid beyond which the inner and outer prongs separate

independent maxillary canal

internarial process of the premaxillary

interparietal

inner prong of the pterygoid

jugal

keeled vomer plate

keeled vomer plate with groove

lacrimal

lacrimal duct

remnant of the lost developing sixth upper postcanine

lingual excavation

lacrimal foramen

recess for the lateral lobe of the cerebellum

large lower foramen of the hypoglossal nerve

lower postcanine number x

replacing lower canine

maxillary antrum

mastoid process

maxillary process of the premaxillary

maxillary

maxillary-palatine suture

nasal

nostril

OPO

OPPT

ORSCRFJ

ORSUSR

THE CRANIAL MORPHOLOGY OF THRINAXODON LIORHINUS SEELEY 399

orbit

opisthotic

outer prong of the pterygoid

opisthotic ridge separating the cochlear recess and the jugular foramen

opisthotic ridge separating utricular and saccular recesses

palatine

prearticular

posterior ampulla

pila antotica

parietal

cultriform process (= rostrum) of the parasphenoid

parietal foramen

pituitary fossa

postorbital

postdentary groove

paroccipital process

paroccipital process abutting against the squamosal

postorbital-parietal suture

posterior palatal foramen

palatal plate of the maxillary

palatal process of the premaxillary

short prong of the palatal process of the premaxillary |

posterior palatine part of the ridge for the attachment of the soft palate

prefrontal

prootic

prootic flange

inner fork of the antero-dorsal process of the prootic

parasphenoid |

parasphenoid-basisphenoid

pterygoid

posttemporal fossa

pterygo-paroccipital foramen

pterygoid process of the quadrate

pterygoid-parasphenoid suture

pterygoid part of the ridge for the attachment of the soft palate

posterior vertical semicircular canal

posterior vertical semicircular canal joins utriculus

premaxillary

quadrate

quadrate canal

quadrate flange

quadratojugal

quadratojugal flange

notch for the flange of the quadratojugal

notch for the flange of the quadrate

quadrate process of the stapes

quadrate ramus of the epipterygoid

quadrate ramus of the pterygoid

ridge for the attachment of the mucous covering of the palate

retroarticular process of the articular

ridge for the attachment of the soft palate

replacing canine

root of the replaced canine

remains of the crown of the replaced canine

root of the erupting replacing canine

root of the functional canine

remaining fragment of the 9th lower postcanine

rhinarial canal

X-th replacing lower postcanine

SMX

ANNALS OF THE SOUTH AFRICAN MUSEUM

root of the third lower incisor

root of the vena capitis lateralis

root of the replaced canine

replacing third upper incisor

X-th replacing upper postcanine

sacculus

surangular

sinus canal

sella turcica

stapedial foramen

septomaxillary

small process at the base of the palatal process of the premaxillary

stapedial process of the quadrate

spur of the prootic

squamosal

flange of the squamosal protruding forward between the quadrate and quadrato-

jugal

saccular recess

splenial

stapes

stapedial recess

supraoccipital

symphysis

tabular

thin coronoid process

transverse flange of the pterygoid

ninth lower postcanine which had not yet erupted

third upper incisor

X-th upper postcanine

upper prong of the palatal process of the premaxillary

utricular recess

unossified side wall of the braincase

utriculus

unossified zone

vomer

vena capitis dorsalis

vena capitis lateralis

venous groove

vomer plate

vena parietalis

wing of the parasphenoid

wall of the recess for the quadrate on the squamosal

wall of the sella turcica

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FIscHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FiscHER, P.-H., DuvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 19605. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika. 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270. ;

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

S. Fourie

THE CRANIAL MORPHOLOGY OF

THRINAXODON LIORHINUS SEELEY

_ VOLUME 65 PART 11 OCTOBER 1974

CAPE ‘TOWN

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 65 ~~ Band

October 1974 Oktober

Parke il- +Deel

ASPECTS OF THE BIOLOGY AND ECOLOGY OF

toe GENUS CYLOS CATREILLE

BRIAN KENSLEY

Cape Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1; 2; 3,,53.7=8); 302, 5, t.=p1.);, SU=2, 5, 79);

6(1, t.-p.i.), 7(1-3), 8, 9(1-2), 10(1),

11(1-2, 5, 7, t.—p.i.), 24(2), 27, 31(1-3), 33

Price of this part/Prys van hierdie deel

R8,70

1974

ISBN 0 949940 51 8

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd., Die Rustica-pers, Edms., Bpk.,

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

ASPECTS OF THE BIOLOGY AND ECOLOGY OF

THE GENUS FYLOS LATREILLE

BRIAN KENSLEY

South African Museum, Cape Town

(With 30 figures and 18 tables)

[MS. accepted 23 August 1973]

CONTENTS

PAGE

Introduction ; : : ‘ ; 401

Description of species : ; : : 402

Geographical distribution : ; ; 412

Habitat : : 2 : 415

Local distribution and burrowing ; 421

Food and feeding

Actual feeding : i ; 431

Food preferences : : 434

Structure of the alimentary canal : 435

Process of digestion ; ; d 44]

Rhythmicity . ; : : 444

Reproduction and erowth: : , 449

Association of 7y/os with other species 451

Comparison of species and discussion . 454

Summary . ; 5 : : : 456

Acknowledgements A ; ‘ 457

References ; : ‘ : % : 457

INTRODUCTION

The isopod genus Ty/os Latreille belongs to the suborder Oniscoidea which

includes a miscellany of more or less terrestrial forms. This group has been split

into two series by Vandel (1943), viz. the Tylienne and the Ligienne. The former

contains two families, the Tylidae and the Stenoniscidae, while the latter contains

about 20 families, including all the true terrestrial isopods. After an exhaustive

investigation into the anatomy and morphology of the Oniscoidea, Vandel

(1943) came to the not surprising conclusion that this heterogeneous suborder

has a polyphyletic origin. The Tylienne series, which is not truly terrestrial and

contains mostly halophilic species which are still dependent on the marine

environment, is derived from the marine Valvifera group, having many structural

similarities. The Ligienne series is derived from some marine family, possibly

the Cirolanidae. This group, which has occupied numerous terrestrial niches,

shows almost perfect adaptation to terrestrial life.

This paper was submitted in 1970 in partial fulfilment of the requirements of the Degree

of Master of Science in Zoology at the University of Stellenbosch.

401

Ann. S. Afr. Mus. 65 (11), 1974: 401-471, 30 figs, 18 tables

402 ANNALS OF THE SOUTH AFRICAN MUSEUM

The family Tylidae contains the single genus 7y/os Latreille comprising at

present 15 known species, the distribution of which may be seen in the following

list, adapted from Vandel (1945).

Southern Group capensis Krauss South Africa

granulatus Krauss South Africa

neozelanicus Chilton New Zealand

spinulosus Dana South America

wegener1 Van Name Venezuela

West Pacific Group granuliferus Budde-Lund Borneo & Japan

opercularis Budde-Lund Philippines

Indian Ocean Group = a/bidus Budde-Lund Nicobar Island

minor Dollfus Seychelles

nudulus Budde-Lund Christmas Island

Red Sea Group exiguus Stebbing Red Sea (possibly synony-

mous with Jatreillei)

Northern Group insularis Van Name Galapagos Island

latreillei Audouin Mediterranean, North and

Central America

niveus Budde-Lund Florida, Cuba, Venezuela

punctatus Holmes & Gray California, Mexico

The similarities of the species of the southern group has led Vandel to suggest

a Gondwana origin for this group, in which he includes the Venezuelan wegeneri

for its similarities to capensis.

Since 1843, when the German naturalist Krauss described T. granulatus and

T. capensis from the Cape, many records of these species have been made. Most of

these merely mention the animals’ occurrence, or are brief morphological notes

with a view to their easy identification. Barnard, in three separate papers (1925,

1934, 1940), has written the most about these species, and while his descriptions

of the structure are excellent, it is felt that some of his ideas regarding the repro-

duction and digestion might require revision.

As both species, by virtue of their large numbers, form an important part

of many sandy beach ecosystems of South Africa, it was felt that an ecological

study would be particularly interesting and would reveal both the similarities

and differences of the two species. This study was thus undertaken in an attempt

to document and explain the many observations made. Simultaneously, the work

on the European T. /atreillei by Mead (1968), Matzakis (1956), Soika (1954), and

Arcangeli (1953), was borne in mind, and the similarities and differences noted.

DESCRIPTION OF SPECIES

As a complete description of the external morphology of both species has

not yet been published, this has been done as a preliminary to the ecological

descriptions. A list as complete as possible of all references to the two species

is included.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 403

Tylos granulatus Krauss, 1843

Tylos granulatus Krauss, 1843: 64, pl. 4, fig. 5. Budde-Lund, 1885: 275; 1906: 74, pl. 3, figs

21-24; 1909: 70. Dollfus, 1895: 352. Stebbing, 1910: 43. Barnard, 1924: 236; 1925: 29,

fig. 3; 1932: 217. Panning, 1924: 172. Vandel, 1945: 227; 1943: 40. Brown, 1959: 470.

Pleistocene fossil: Haughton, 1931: 27.

Body oval—elongate, widest at fifth or sixth segment. Integument chiti-

nous, transparent, thick, covered with numerous flattened roughly elliptical

granules (Fig. 4c). Membranous subintegumental layer bearing branched chro-

matophores. Head (Fig. 1) evenly convex, bearing two compound eyes dorso-

vertex

first antenna

second antenna

frontal process

clypeus

labrum

gena

mandible

first maxilla

second maxilla

maxilliped

Fig. 1. Ty/os granulatus. Anterior view of head.

laterally, each composed of about 50 ommatidia. No distinct frontal line, but

faint impressed line joining anterior margins of eyes, becoming obsolete mid-

dorsally. Vertex narrowed antero-ventrally between first antennae, giving way to

broadly rounded frontal process. Broad clypeus below frontal process, with

labrum ventrally attached. Labrum asymmetrical due to underlying left mandi-

ble atop right mandible. Narrow gena situated lateral to second antennae, ven-

trally embracing mandibular bases.

Head appendages—first antenna consisting of single almost immobile

segment medio-dorsal to second antenna, dorsally flattened, triangular, level

with rest of vertex, ventrally with broad articulating area, with socket-like

depression near apex of triangle, containing numerous aesthetascs. Second

antenna (Fig. 2c) stout, elongate, consisting of five basal segments plus flagellum

404 ANNALS OF THE SOUTH AFRICAN MUSEUM

| ae eee re st Pd

0 5mm

Fig. 2. a. Tylos granulatus. External view of first antenna. b. Internal view of first antenna.

c. Second antenna. d. 7y/os capensis. External view of first antenna. e. Internal view of first

antenna. f. Second antenna.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 405

of four smaller segments, terminal segment minute. Surface granular, with

numerous setae, especially on terminal and subterminal segments. Mandibles

powerful, heavily chitinised. Left mandible (Fig. 3a) with incisor process of

three powerful teeth. Lacinia mobilis also strongly chitinized with three teeth,

y/ |

ee | e

Fig. 3. Tylos granulatus. a. Left mandible. b. Right mandible. c. First maxilla. d. Second maxilla.

e. Maxilliped. f. Labium.

406 ANNALS OF THE SOUTH AFRICAN MUSEUM

with setose lobe at base. Latter followed by group of 10-12 penicillae, exterior

to which, heavily setose molar process bearing terminal rasp-like process. Right

mandible (Fig. 3b) similar to left, differing only in lacinia mobilis, which is very

reduced, not heavily chitinized. Incisor process fits between lacinia mobilis and

incisor process of left mandible. First maxilla (Fig. 3c) bilobed, consisting of

two endites, outer lobe longer than inner, with setae along entire outer margin,

bearing about 11 or 12 terminal hooked teeth of varying size, larger ones strongly

chitinised. Inner lobe terminated with three stout setose penicillae. Second

maxilla (Fig. 3d) consisting of single broad lamella, terminally heavily setose,

bearing groove on inner face. Maxilliped (Fig. 3e) consisting of coxa, and baso-

podite bearing palp and endite. Broad strong palp with most of external surface

armed with flattened blunt spines. Median edge divisible into three sections,

reflecting fusion of three segments, each bearing numerous close-set blunt spines.

Endite smaller than palp, terminated with five setose penicillae. Lobed lower lip

(Fig. 3f) just ventral to mandibles, heavily setose on medio-dorsal surfaces.

Tongue-like median setose lobe at base of outer lobes.

Pereion—consisting of seven segments, all except first having epimerites

(Fig. 4a), those of segments six and seven being quadrate, large, those of other

PCI Se

Fig. 4. a. Tylos granulatus. Pereion epimerites.

b. Tylos capensis. Pereion epimerites. c. One

square millimetre of exoskeleton surface, showing

granulation.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 407

segments being smaller, rectangular/triangular. Seven pairs of pereiopods

present, first four pairs being forwardly directed, differing slightly from three

posteriorly directed pairs. Each pereiopod consisting of six segments, basopodite

largest. An obvious angle between basis and rest of leg. Anterior four pairs of

pereiopods (Fig. 5a) with basis equal to ischium + carpus + merus in length.

Merus with large rounded dorsal lobe. Carpus, propodus, and dactylus together

forming a claw. Numerous spines and setae present on all segments, especially

on more terminal ones. Posterior three pairs of pereiopods (Fig. 5b) with basis

equal to ischium + merus in length. Dactylus small; propodus, carpus, and

merus stout, with strong spination. In ovigerous females, lamellar oostegites

present on all pereiopods.

Fig. 5. Tylos granulatus. a. Second pereiopod. b. Fifth pereiopod.

Pleon—consisting of five free segments, plus rectangular telson. Latter

twice broader than long. Five pairs pleopods present (Fig. 6), first pair reduced

to slender lamella, posterior four pairs consisting of basis plus endo- and

exopodite. Both latter lamellar, performing respiratory function. Exopod with

numerous vertical folds, endopod with irregular folds. Second to fourth exopods

with posterior pointed process medially. Endopod of second pleopod in male

408 ANNALS OF THE SOUTH AFRICAN MUSEUM

210m

2mm

Fig. 6. Tylos granulatus. Pleopods one to five, male. (Two to five drawn to scale.)

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 409

modified to form long copulatory stylet medially, stretching posteriorly to level

of fourth pleopods. Uropods valve-like, ventral, covering anal aperture, con-

sisting of flattened basal portion with tiny setose endopod posteriorly. No sign

of bilamellar uropod as mentioned by Stebbing (1910). Ventrally, segments

three to five of pleon forming broad plates, those of segment five being antero-

medially produced around the uropods, almost meeting in midline (Fig. 7a).

second pleopod

last pereiopod

i base

fifth pleon

plate

uropod

telson

Fig. 7. Ventral view of pleon. a. Tylos granulatus. b. Tylos capensis.

410 ANNALS OF THE SOUTH AFRICAN MUSEUM

Tylos capensis Krauss, 1843

Tylos capensis Krauss, 1843: 64, pl. 4, fig. 6. Budde-Lund, 1885: 276; 1906: 73. Dollfus, 1895:

352. Stebbing, 1910: 43. Barnard, 1932: 218, pl. 3, figs 14-18. Vandel, 1945: 227; 1952: 192.

Tylos granulatus (non Krauss), Collinge, 1945: 345.

Tylos incurvus Budde-Lund, 1906: 79, pl. 3, fig. 41.

Body oval-elongate, widest at fifth pereion segment. Integument chitinous,

transparent, thick, minutely setulose, smooth. Membranous subintegumental

layer bearing numerous branched chromatophores.

Head—evenly convex, bearing two compound eyes dorso-laterally, each

composed of about 40 ommatidia. No distinct frontal line, but faint impressed

line joining anterior margins of eyes, becoming obsolete medio-dorsally. Antero-

ventrally the vertex narrowing between first antennae, giving way to broadly

rounded frontal process, ventral to which, the broad clypeus, with labrum

attached ventrally to it. Labrum not as asymmetrical as in T. granulatus. Narrow

lateral genae embrace mandibular bases ventrally.

Head appendages —first antenna (Fig. 2d-e) consisting of single almost

immobile segment, medio-dorsal to second antenna, dorsally flattened, roughly

triangular, level with vertex surface, ventrally with broad articulating area,

having socket-like depression near apex, containing numerous aesthetascs.

Second antenna (Fig. 2f) stout, elongate, consisting of five basal segments plus

flagellum of four smaller segments, terminal segment being proportionally

larger than in 7. granulatus. Second basal segment also differs from 7. granulatus.

Entire appendage setose. Mandibles powerful, heavily chitinized, left mandible

(Fig. 8a) with incisor process consisting of three strong teeth, lacinia mobilis

also of three powerful chitinous teeth, with setose lobe at base, followed by

group of 10-12 penicillae, external to which, heavily setose molar process,

terminated with rasp-like process. Right mandible (Fig. 8b) similar to left,

differing only in lacinia mobilis, reduced, not heavily chitinized. Incisor process

fits between lacinia mobilis and incisor process of left mandible. Labium bilobed

and heavily setose, situated ventral to mandible, with median tongue-like

setose lobe. First maxilla (Fig. 8c) bilobed, consisting of two endites. Outer

lobe longer than inner, bearing setae on outer edge, with 10-12 curved terminal

teeth of varying size, the larger teeth being heavily chitinized. Inner lobe ter-

minated with three stout setose penicillae. Second maxilla (Fig. 8d) consisting

of weakly chitinized broad plate, terminally heavily setose, with groove on

inner face. Maxillipeds (Fig. 8e) meeting in mid-line, forming efficient lower lip,

consisting of coxa and basopodite bearing broad palp and more slender endite.

Palp with external surface scattered with blunt spines. Median edge divisible

into three sections, reflecting fusion of three segments, each bearing numerous

paddle-shaped modified spines. Endite smaller than palp, terminated with five

setose penicillae.

Pereion—consisting of seven free segments, all except first bearing epi-

merites (Fig. 4b), those of segments six and seven quadrate, those of other seg-

ments smaller, ventrally rounded. Seven pairs of pereiopods present, first four

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 411

Fig. 8. Tylos capensis. a. Left mandible. b. Right mandible. c. First maxilla. d. Second maxilla.

e. Maxilliped.

pairs forwardly directed, remaining three pairs posteriorly directed. Each

pereiopod consisting of six segments, basopodite largest. Obvious angle between

basis and rest of leg. Anterior four pairs with basis equal to merus + ischium

+ carpus in length. Merus with large dorsally rounded lobe. Carpus, propodus,

412 ANNALS OF THE SOUTH AFRICAN MUSEUM

and dactyl together forming a claw. Numerous spines and setae on all segments,

especially on more terminal ones. Posterior three pairs with basis equal to

ischium -++ merus in length. Dactyl small; propodus, carpus, and merus stout,

with strong spination.

Pleon—consisting of five free segments plus rectangular telson, latter

broader than long. Five pairs of pleopods present, first reduced to slender

lamella. Posterior four pairs consisting of basis, endo- and exopodite. Latter

lamellar with numerous vertical slits, unlike 7. granulatus, which has numerous

folds. Endopods with irregular folds. Endopod of second pleopod of male

modified to form median copulatory stylet, stretching posteriorly to level of

fourth pleopod (Fig. 9). Uropods valve-like, ventral, covering anal aperture,

consisting of single flattened lamina, with thickened ring-like inner portion,

bearing tiny setose endopod posteriorly. Ventrally, fifth segment of pleon

not produced antero-medially as in 7. granulatus (Fig. 7b).

CULL

Fig. 9. Tylos capensis. First and second pleopods, male.

GEOGRAPHICAL DISTRIBUTION

Tylos granulatus

Krauss (1843) recorded his new species of 7. granulatus from Table Bay.

On the Cape Peninsula this species has been collected by the author at Hout Bay,

Noordhoek, Olifantsbosch, Schuster’s River mouth, and Witsandsbaai. Intensive

collecting shows that the southern limit of the species is about 24 km from Cape

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 413

Point, and about 48-56 km from the first record of T. capensis, viz. Simon’s Bay.

Collecting along the coast of South West Africa has established the northern

limit of the species to be in the region of the Ventura wreck site (19.10S., 12.37E.).

This record extends the known range of the species by 480 km, the most northerly

published record being Swakopmund (Panning 1924). Although the coast

from this region north to the Kunene River was covered, as well as the southern

coast of Angola, no further specimens were found. 7. granulatus thus has a

range of about | 600 km. It was noted that towards the end of the northerly

range the animals were not found on open beaches but were burrowing in the

sand between rocks. A possible explanation for this was found in one of the

limiting factors which prevent the northerly spread of the species. This is the

ghost crab Ocypode cursor which inhabits sandy beaches and burrows into sand.

This crab which occurs in large numbers, feeds voraciously on any organic

material thrown ashore. Like the fiddler crabs of the same family, O. cursor

seems to possess a tidal rhythm, and was observed to be active by day and at

night. There can be little doubt that the presence of this crab presents a strong

barrier to the isopods, feeding on the material which would normally form the

food of Tylos. O. cursor occurs in the Mediterranean and along the entire west

African coast. The southernmost record is 6,5 km north of Méwe Point, South

West Africa. (19.23S., 12.42E.). The overlap of the two species is thus a matter of

about 21 km, in which area neither is very abundant.

Haughton (1931), during a geological survey of the formations of the west

coast, collected a fossil isopod from the oyster line of the diamondiferous deposit

of Alexander Bay. The isopod was tentatively identified as 7. granulatus. The

hard matrix which concealed the ventral surface of the animal has now been

drilled away. The granular surface together with the form of the ventral process

of the fifth pereion segment establishes the first identification as being correct.

The age of the oyster line is placed as mid Pleistocene. The species thus has a

minimal age of 500 000 years. Taking rates of evolution into consideration it is

probable that this species was already established at the start of the Pleistocene.

Tylos capensis

This species was first recorded by Krauss in 1843 from Table Bay, but this

must have resulted from an error in labelling, as no record of T. capensis from

the west coast has since been established. The species occurs eastward from

False Bay. Budde-Lund (1906) records it from Simon’s Bay, but this population

has died out, perhaps due to naval and municipal development in the area.

From Muizenberg eastward, the species occurs at most of the open beaches,

including Strandfontein, Macassar Beach, Somerset Strand, Gordon’s Bay,

but is nowhere very plentiful. Along the south and east coasts the species occurs

intermittently and is common in the Port Beaufort and Knysna areas. North of

Port Elizabeth the occurrence could not be confirmed, but probably stops

somewhere on the southern Natal coast. Barnard (1932: 219) makes the follow-

ing statement: ‘These animals form one of the most marked differences between

414 ANNALS OF THE SOUTH AFRICAN MUSEUM

the faunas of the west and the east sides of the Cape Peninsula. If this separation

of the two species, one from Table Bay northward and the other from False Bay

eastward is proved to be a fact by further and more intensive collecting, it leads

to the interesting though perhaps fruitless speculation as to why there was no

transgression of the one species into the area of the other, when the sea was

continuous across the present Cape Flats between Table Bay and False Bay.’

Method of dispersion

The actual method of dispersion of the species is still under dispute. Two

main possibilities exist, viz. by sea or over land. Sea dispersion does not seem

at all impossible. The survival of various sizes of T. granulatus in sea water was

tested. The results are given in the following table.

(Temperature of water —17°C. Juveniles—under 20 mm length)

Time

Adults 1900

Juveniles 2000

0900

(next day) (next day)

Although by no means conclusive, it would appear that the juveniles have a

greater survival time than the adults in sea water. The death of the adults is

probably due to oxygen starvation, as the area of the pleopods is insufficient to

take enough oxygen from the water by diffusion. The pleopod area in the

juveniles is proportionally much greater than in the adults. From observations

on beaches, it was seen that juveniles tend to feed lower on the beach than adults.

The juveniles are usually found on the debris lines, the adults only occasionally.

On the debris lines there is the danger of powerful waves swamping the feeders.

This was often seen at Blouberg (an area where intensive observations were

carried out). The juveniles roll up into a ball when exposed to an incoming surf.

This rolling encloses a bubble of air between the setose pereiopods, making them

buoyant. The result of being caught in the surf is that they are deposited on the

beach at the new debris line, along with pieces of food. This would seem to be a

useful adaptation for food-finding. It is possible that with a very powerful wave,

instead of being left behind on the sand, the juveniles are swept to sea. A survival

of 12 hours at sea would give them a good chance of being thrown up on an

adjacent beach. This would explain the distribution of the species in such places

as Liideritzbucht, where small sandy bays alternate with steep rocky outcrops,

over which it would be impossible for the animals to climb. It is interesting to

note that Menzies (1952) records having taken two almost adult specimens of

Tylos punctatus in a surface plankton haul in Newport Harbour, California.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 415

Menzies notes that T. punctatus has a very discontinuous distribution along the

coast, being found around bays and estuaries, and never on open coasts. He

concludes that this record may strengthen the hypothesis that such discontinuous

distribution can be accounted for by the animals being carried by ocean currents.

The behaviour of adults caught in the surf is interesting. As the incoming

surf sweeps over them, they ‘flatten out’, allowing the water to wash over them

and then back. As soon as the water has receded, they can be seen to move

rapidly up the beach, out of the reach of the waves.

Land dispersion 1s almost certainly used under certain conditions. The most

obvious example is that of the Orange River mouth. Brown (1959) noted that

when the river mouth was open to the sea 7. granulatus was to be found in the

sand on the southern bank. In March 1968 the mouth was completely blocked

by an extensive sand bar, along the entire length of which the isopods were to be

found. This sand bar had been in existence for two years since the last flooding of

the river. This latter event occurs spasmodically, sometimes consecutively for

several years, or more usually once every two or three years. Between floodings

the sand bar is built up. Within two years, Ty/os had thus colonized a sand bar

about 3 km long. The most obvious method of colonization would be by moving

in from the undisturbed banks of the mouth. Since the Orange River population

is to be found on the landward side of a crest on the beach, it is unlikely that the

alternate method of colonization occurs here.

HABITAT

Both Tylos granulatus and T. capensis inhabit sandy beaches of the coasts

of southern Africa. On these beaches the animals are exposed to numerous

environmental factors, the influence of each varying in importance to the ani-

mals. These factors include the sand itself, the sea and its tides, the water table

in the sand, temperature fluctuations, humidity, light, precipitation, in addition

to several extraneous factors. Although some of these factors are discussed

separately below, it must be borne in mind that they all interact to produce the

overall habitat.

TYPE OF BEACH

Tylos granulatus

After numerous observations, it can be said that this species frequents.

beaches which almost invariably are exposed to strong wave action, uninter-

rupted by rocks. This type of beach is common along the west coast, all the

localities listed in the section on distribution being of this type. The preference

for the exposed beach could clearly be seen in several places, particularly in

Griffith’s Bay, Liideritzbucht, where there is a small exposed beach, as well as a

long stretch of low rocks. Behind the latter in the HWS to HWN region, is a

band of sand similar to that of the exposed portion. Tylos was found on the

exposed part, up to the edge of the rocks but not in the sand behind the rocks.

416 ANNALS OF THE SOUTH AFRICAN MUSEUM

The slope of the beach would also appear to be of some importance. The

populations are spread out on flat beaches, where the high tide floods a large

area. This was clearly seen at Groenriviermond and Stormvogelbucht. On

beaches with a moderate gradient such as Blouberg (slope of 1/8) the population

is still fairly outspread. At the mouth of the Orange River, where the population

inhabits the sand bar at the actual mouth of the river, the beach has a slope

of 1/4, and the animals are found on the landward side of the crest. The latter is

at the HWN mark. Presumably to live on the seaward side of the crest would

expose the animals to too much wave action.

Tylos capensis

Unlike T. granulatus, this species does not seem to be as selective in its

choice of beach, being found on exposed as well as on sheltered beaches. At

Strandfontein, an exposed beach with a very gentle slope (about 1/20), the

population is spread out along the HWS region. At the mouth of the Breé River

the population is found on a short steep beach (1/6) within the actual mouth of

the river. Although quite sheltered, there is still some wave action, mostly due

to wind on the open expanse of water. At Knysna the population is found at the

lower reaches of the lagoon estuary, again on a short steep beach, but where the

wave action was very slight.

SAND

Tylos granulatus

The actual type of sand in which Ty/os burrows is very variable, and it

would seem that the animal has a wide tolerance of sand grain size. At Griffith's

Bay and Groenriviermond the sand is very fine, forming a soft mud when

saturated with water. The Blouberg sand varies from fairly fine white quartz

grains to layers of coarser mixed grains and shell fragments, very compacted.

The sand at the mouth of the Orange River is dark brown, consisting of grains of

garnet, amethyst, granite, fairly coarse and very friable. At Blouberg the area

inhabited by Ty/os ends with the start of a large reef of rock. In this area the

beach consists of fine pebbles all about 2-3 mm in diameter. No animals are to

be found here, but 20 animals transferred to this area all burrowed to a depth

of about 120 mm. The most unusual substrate in which this species was found

was observed at Griffith’s Bay, where the upper 127 mm consisted of fine white

sand. This gradually gave way to a 76 mm layer of grit and small pebbles. Below

this was a 152 mm layer of limpet and mytilid shells, followed by large rocks

and shells to a depth of 610 mm. Animals are found in this very coarse substrate

to a depth of about 305 mm.

Tylos capensis

As in T. granulatus this species has been found in a variety of sand types,

viz. very fine quartz sand with shell fragments at Strandfontein, friable fairly

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 417

fine sand at the mouth of the Breé River, and uniformly fine sand at Knysna.

T. latreillei from the Mediterranean and North American coasts appears to

be very different from both South African species in its choice of habitat. Arcan-

geli (1953) remarked that 7. /atreillei avoided beaches of fine sand, nor was it

found at river mouths where the sand tended to mud, as its pereiopods impaired

its progress. Instead, it inhabited crevices of rocks, or lived amongst stones and

pebbles, or sometimes in coarse sand. 7. punctatus of the North American west

coast also differs from the southern African species in choice of habitat. Menzies

(1952) notes that T. punctatus inhabits sandy beaches, but is found under logs

and debris where the sand is usually dry and seldom inundated by the high tides.

TIDES AND THE WATER TABLE

Tylos capensis and T. granulatus

The importance of the tides is more fully discussed in the section dealing

with rhythmicity. Arcangeli (1953) in his discussion of the ecology of T. /atreillei

remarks that this species is ‘indifferent’ to the tides—quite the opposite to the

present species. The water table varies with the tides but is unlikely to have any

direct effect on the animals. From Figure 17 of the transect of the Blouberg

area studied, it can be seen that the water table is too far below the level to

which the animals burrow to influence them.

WIND AND PRECIPITATION

Tylos capensis and T. granulatus

Light rain and mist have no apparent effect on the animals, while gale

force winds do not seem to interfere with their feeding activities.

TEMPERATURE

Tylos capensis and T. granulatus

As can be seen from Figure 10, the temperature within the sand at the level

of the high tide mark is subject to less fluctuation than the surface sand. As

under normal conditions no extremes of temperature are experienced, it seems

unlikely that the daily temperature changes have any influence on the animals.

Seasonal temperature changes are of greater importance and are probably

involved in the timing of the reproductive cycle.

HUMIDITY

Tylos capensis and T. granulatus

As Tylos breathes by diffusion of gases through modified pleopods which are

exposed on the ventral portion of the pleon, humidity is of great importance.

The habits of the animal thus preclude any serious exposure to desiccating

conditions. The sand, even a few metres above the HWS mark, has a relative

418 ANNALS OF THE SOUTH AFRICAN MUSEUM

hours

Fig. 10. Temperature changes over 24 hours in midsummer at Blouberg.

temperatunes a. oe temperature at 300 mm. — — — temperature at 450 mm.

surface

humidity of more than 90°% at 75 mm below the surface. The upper 75 mm of

sand is frequently dried out during the day by sun and wind, but as no animals

are in this layer during the daylight hours, this presents no danger. At night

with the drop in temperature, spray from the sea together with any dew deposi-

tion keeps the air close to the sand almost saturated with water vapour, thereby

preventing desiccation.

LIGHT

Tylos capensis and T. granulatus

It is not possible for sunlight to penetrate more than 25 mm below the

surface of the sand; sunlight would thus have very little direct influence on the

animals. When exposed to sunlight, the animals rapidly burrow into the sand,

and can be said to be strongly negatively photosensitive. When not feeding,

both species react to both white and coloured light by becoming immobile,

but bright white light has little immediate effect on feeding animals.

EXTRANEOUS FACTORS

During week-ends, both areas studied form part of a popular beach, and

are often continuously trampled. This does not seem to have any effect on the

nightly emergence of the animals.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 419

Oil pollution

This was noted at the Blouberg population studied, when several observa-

tions were made. Crude oil from the ballast tanks of a tanker washed ashore on

7 February 1970. The following day was a spring tide, the result being that the

beach was covered with oil from the water line to the HWS mark (Fig. Ila). At

Fig. 11. a. Oil pollution at Blouberg. b. Exit holes through thick oil.

the lower levels wave action tended to froth the oil, which was therefore not

very thick. Higher up the beach the oil formed a thick solid deposit. Low tide

on the night of the 7th was at 2225 hours, when not many animals emerged.

Almost all the animals which did emerge were killed, death probably being

caused by clogging of the pleopods and a consequent arrest of respiration. About

420 ANNALS OF THE SOUTH AFRICAN MUSEUM

30 dead animals and about the same number of exit holes were found. On the

night of the 8th, low tide was at 2303 hours, when numerous animals emerged

(Fig. 11b). Where the oil was not very thick, the animals returned at the high

water mark (Fig. 12a). Where the oil formed a thick layer at the high water

Fig 12. a. Return mounds on thin oil layer. b. Return mounds

amongst vegetation in low sand dunes, well above HWS.

mark, a few animals returned at a lower level. Many moved above the high

water mark, into the low sand dunes, and burrowed amongst the vegetation

(Fig. 12b), something not seen before. Of the animals which entered the sand

through a thin layer of oil, several were dug up. No trace of oil on the bodies

other than at the tips of the pereiopods between the bristles could be seen.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 421

Presumably the corkscrew action used in burrowing rubbed off any adhering oil.

A food problem was posed as almost all the seaweed was oil-covered. Above

the high water mark, very little vegetable matter was present other than the

actual dune plants, some of which were eaten. Humidity did not pose any prob-

lems, in spite of the fact that the dry surface layer of sand was deeper than at the

high water mark. After about three weeks the animals returned to their usual

level on the beach, as some of the oil had been removed, while the thinner layers

had been covered by wind-blown sand.

LOCAL DISTRIBUTION AND BURROWING

The local distribution of TJy/os may be discussed under three headings, viz.

distribution along the beach, 1.e. parallel with the sea; distribution on the beach

at right angles to the sea; vertical distribution in the sand. The actual distribution

of Ty/os on any section of beach can be judged by the spread of either the exit

holes or the return mounds. This judging of the population spread is easy at the

time of spring tides, when the animals emerge and return on clean-swept sand.

At the time of neap tides, however, the sea does not remove each day’s mounds

and holes. An overlapping then occurs, obscuring the distribution. Strong winds

also flatten mounds and fill holes, obscuring the spread. A more definite check

can be made by digging a trench up the beach, or by actually observing the

animals emerging at night.

DISTRIBUTION ALONG THE BEACH

Tylos granulatus

The most important fact to emerge from prolonged observation of a

population, is that it is, as a whole, clustered mainly around the high water mark.

An explanation for this is that food is deposited at the high tide mark. During

spring tides, algae which have been torn loose are left far up the beach. The

Tylos population at high water of springs is thus far up the beach. At the same

time, if a long stretch of beach is observed (about 500 metres) it can be seen

that the population is split up into separate clusters along the HWS mark

(Fig. 13a). As the height of each tide decreases after the spring tide, the popula-

tion spreads out, some individuals following the high tide mark as it progresses

down the beach, others remaining higher up. The result is that whereas at spring

tide the population is split into clusters, at neaps it is outspread and forms a

continuous band along the beach (Fig. 13b).

Tylos capensis

Although never as abundant as 7. granulatus, this species shows a similar

distribution along the beach, being clustered high up the beach at springs, and

very spread out at neaps.

422 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 13. a. A cluster of return mounds of 7. granulatus at HWS, Blouberg. b. Blouberg at

neaps. Exit holes at left. High tide line at centre marked by blocks of wood, etc. Return mounds

at right.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 423

DISTRIBUTION ON THE BEACH AT RIGHT ANGLES TO THE SEA

Tylos granulatus

The term ‘at right angles to the sea’ is used in preference to ‘vertical zona-

tion’ as in this case the situation is complicated by the additional distribution

vertically in the sand, and by the change in spread with the change in the height

of the high water mark. The spread of the population up the beach appears to be

related to the range of the high tide. On a beach which has a very flat high water

area, such as Groenriviermond or Stormvogelbucht, the population is spread

over an area about 100 metres in width. At Blouberg where the beach has a

slope of about 1/8, the population inhabits an area about 20 metres in width.

Figure 14 indicates the concentration of animals in two half square metres. Set

(a), 10 metres above HWN, contained five animals; set (b) at HWN, contained

60 animals. Ten metres below HWN no animals were present.

The Orange River population is unusual in that it is situated on the land-

ward side of a sand ridge which slopes towards the river pools. Here the area

occupied is about 30 metres in extent.

Tylos capensis

A similar situation prevails with this species. At Strandfontein, with a

slope of about 1/20, the population is spread over about 40 metres. The Breé

River population on a much steeper slope occupies an area with a width of

about 10 metres.

VERTICAL DISTRIBUTION IN THE SAND

This was obtained by excavating random half square metres. (This area

was preferred to that of a full square metre, due to the time needed to complete

an excavation.) The volume of sand under this half square metre was divided

into layers each 80 mm deep. The distribution of animals in each 80 mm layer

was then plotted on rectangles. This excavation was done in each case until no

further animals were encountered. The time of day at which the excavation was

done was recorded. These resuits may be seen in Figures 14 to 16.

Tylos granulatus

In Figure 14a—c it can be seen that in the late afternoon the majority of

animals are between 80 and 240 mm below the surface. At sunset, the majority.

are distributed from just below the surface to about 160 mm down. At Yster-

fontein (Fig. 14), a beach similar to Blouberg, at about midday the majority

are from 240 to 320 mm below the surface. Figure 15a—c of the Orange River

mouth shows the distribution of the animals over a length of about 30 metres of

beach. The positions of the half square metres may be seen on Figure 17a. The

juveniles appear to be concentrated in the upper layers of the sand. As at Yster-

fontein and Blouberg at midday, the adults are concentrated at a depth of

160 to 320 mm below the surface. This is again seen in Figure 16d of Agate

424 ANNALS OF THE SOUTH AFRICAN MUSEUM

depth (cm)

0-8

8 -16

16 - 24

24- 32

32-40

a b

1700 hrs 1700 hrs

S16

16 -24

24-32

32-40

2000 hrs 1330 hrs

Fig. 14. Vertical distributoin of Ty/os granulatus. a—c. Blouberg. d. Ysterfontein.

8 — 16

16-24

24-32

32-40

3)— 16

16-24

24-32

32-40

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE

Fig. 15. Vertical distribution of Tylos granulatus. a-c. Orange River mouth sand bar. d. Agate

Beach, Liideritzbucht. e. Griffith’s Bay, Liideritzbucht.

426 ANNALS OF THE SOUTH AFRICAN MUSEUM

depth (cm

a 46

16-24

24-32

O= 6

3 = 15

16 — 24

24-32

Fig. 16. Vertical distribution of Tylos capensis. a—c. Strandfontein. d-e. Breé River.

Beach, Liideritzbucht, where the juveniles are found in the upper layers. Figure

15e of Griffith’s Bay, Liideritzbucht, represents an unusual vertical distribution

which can only be satisfactorily explained by reference to the substrate in which

the isopods were found. From the figure it can be seen that whereas the adults

are limited to the depths of 80 to 240 mm, the juveniles, although concentrated

in the upper 80 mm, are spread throughout the sand to a depth of 320 mm.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 427

For midday the adults are unusually close to the surface, but they could not have

burrowed deeper. The fine sand penetrated to a depth of 120 mm but gradually

gave way to grit and small pebbles to about 200 mm. This layer was replaced

by a deeper layer of large stones and shells. The juveniles could move between

these large objects, but not the adults.

Tylos capensis

The vertical distribution of this species was also investigated by means of

half square metre excavations, at both Strandfontein and the Breé River

(Fig. 16a—e). Beach profiles may be seen in Figure 17c—d, on which the positions

of half square metres are indicated by the letters ‘a, b, c.’ While at no time

abundant, it would seem that this species seldom burrows to a depth of more

than about 200 mm.

METHOD OF BURROWING

Burrowing was observed on the beach as well as in the laboratory. Captive

animals were observed in large perspex/glassfibre containers, as well as in

limoria. These latter being 30 mm wide, just allow an adult 7y/os space to turn a

complete circle when digging almost vertically downwards. This meant that the

animals were always visible from either of the glass sides at any one point. All

observations were done in the dark, using a torch at brief intervals, so as to

disturb the animals as little as possible. The sequence of actions used in burrow-

ing is as follows: the animal scoops out a hollow in the sand, using the anterior

three pairs of pereiopods. These pass the sand on to the posterior four pairs of

pereiopods, which push the sand backwards, to form a little heap behind the

animal. The animal then moves sideways, through an angle of about 45°,

digs out more sand with the first three pairs of pereiopods, pushes it back with

the posterior legs, and forms another heap of sand next to the previous one.

This sequence continues, usually five to seven pushes of sand, followed by a

pause of about a minute. Each short burst of digging takes the animal a little

deeper into the sand, as it moves laterally (usually anti-clockwise) after each

push. The overall movement is a helix-shaped path passing vertically through

the sand. One complete turn of the helix represents eight or nine backward

pushes of sand, and takes the animal about 30 mm into the sand. On the surface

this complete turn can be seen as an irregular cone of sand, having eight or nine

small peaks if the sand is moist and the grains clinging. When the animal starts

digging, the antero-posterior plane is horizontal, but as it progresses, this plane

tilts upwards, so that once the animal is in the sand, this plane is vertical. Once

below the surface, the backward pushing of the posterior pereiopods closes the

hole. As the sand is usually moist below the surface the backward pushing tamps

the sand, preventing it from falling back and hindering progress. The actual

speed of digging varies, depending on the number of rests. Timing of six animals

shows that the usual speed is about 200 mm/12-15 minutes. Once the animal

has reached the ‘desired’ level, it moves around, flexing the pereiopods. This

428 ANNALS OF THE SOUTH AFRICAN MUSEUM

river, pools C b a LWS

a. sand bar at Orange River mouth

HWS a ? C LWS

HWN

b. Bloube rg

a b C LWS

HWS

C. Strandfontein

LWS

d. Breérivier

Da ee

scale: 1 div = 10m

Fig. 17. Beach profiles. LWS—Low water of springs. HWN-—High water of neaps. HWS —

High water of springs. a. Orange River mouth sand bar. b. Blouberg. c. Strandfontein. d. Breé

River.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 429

causes its rounded dorsal surface to press against the sand, thus forming a

chamber only slightly larger than itself. Once this is done, the animal curls up

and remains in this position. Usually with the eyes ventral, until the time of

emerging approaches.

It was sometimes seen that an animal will start digging on a mound, i.e.

where another animal has gone down. The surface counts of mounds thus need

not give an accurate indication of the number of animals below the surface.

For example, at Lambert’s Bay the surface count for a random half square

metre was 55 mounds, but excavation yielded about 120 adults. Although several

animals may enter the sand at the same point, once in the sand they separate

and emerge from separate holes.

The actual time taken to emerge, once the animal becomes active, is not

known, but is probably much longer than the return digging. This is inferred

from observations made at different times of the day. For example, at 1700

hours most of the adults are 80 to 240 mm below the surface. At 2000 hours at

the same locality, most of the adults are from just below the surface to 160 mm

below. The animals may dig upward either following the path made by their

entry or may follow a fresh path. Once the animal reaches the surface and the

eyes are exposed, it pauses (Fig. 18a). This pause was observed in all emerging

animals, varied in duration, but was usually between 2-15 minutes. Once the

animal has left the burrow, its point of exit is marked by a circular hole up to

30 mm deep (Fig. 18b). The animal then immediately starts a rapid apparently

random perambulation, stopping only when food is encountered.

From this study of the local distribution and burrowing of both species of

Tylos the following facts emerge:

1. Tylos usually burrows close to the high water mark.

2. At HWS the population is high on the beach and clustered.

3. At HWN the population is lower on the beach and outspread.

4. The animals are distributed in the sand up to a depth of 500 mm in the case

of T. granulatus, 200 mm in T. capensis.

The animals burrow with a spiral movement.

Re-entry into the sand leaves a cone-shaped mound on the surface.

7. The speed of return to the surface is slower than the speed of re-entry into

the sand.

8. Exit points are marked by a shallow circular hole.

9. The animals sometimes return via a path made by another animal.

10. All digging is approximately vertical.

FOOD AND FEEDING

The section dealing with feeding and related topics in Ty/os has been divided

into several sections, viz. actual feeding, i.e. the obtaining of the food, the food

preferences, the structure of the alimentary canal, and the actual process of

digestion.

430 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 18. a. 7. granulatus. Animal at left starting to burrow, animal at right just emerging. Exit

holes at upper right. b. Numerous exit holes of 7. granulatus.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 431

ACTUAL FEEDING

The feeding of both 7. capensis and T. granulatus is confined to the brief

period (two to three hours) that the animals are emerged. It was found that

Tylos tends to burrow around the high tide mark, moving down the beach as

the height of successive tides drops, or up the beach as the level approaches that

of spring tide. This is a behavioural adaptation connected with food supply, as

any floating debris including sea weeds and dead animals will be left on or near

the high tide mark. The mouthparts are adapted for rapidly cutting pieces of

food. The gut, particularly the midgut portion, is very capacious. Both these

factors allow the animal to eat a maximum amount in the time available. A very

rough measure of the amount eaten in the period emerged may be obtained from

the following data (see Table 3). Ten specimens of 7. granulatus were captured

as they emerged to feed, weighed, marked, and allowed to feed. These animals

were again captured at the end of the feeding period, reweighed and their length

measured. They were then killed and the volume of the food in the gut measured

by displacement. From this it was found that the animals ate up to 25°%% of

their body weight in the two-hour period involved. A similar situation was found

to apply in T. capensis.

From observations of animals feeding, a list of foodstuffs was compiled,

which is summarized in Table 4. From a survey of the food consumed by the

isopods it can be seen that both species are omnivorous, but with a bias towards

an herbivorous diet. In this, 7. granulatus and T. capensis resemble the European

T. latreillei which Arcangeli (1953) describes as feeding on ‘decomposed

matter’.

From observation it was seen that the majority of adults feed almost

exclusively on algae, while the juveniles (less than 15 mm in length) feed mainly

on animal matter. The adults which are found around the high water mark feed

mainly on the larger brown algae thrown ashore, along with the epiphytes on

these algae. It would seem from the table of food matter, that the red algae form

the major portion of the food, but relative sizes must be borne in mind. Whereas

most of the red algae are small flattened plants, seldom more than 100 mm in

length, the brown algae are often several metres in length, with massive stipes.

All three brown algae recorded are large plants.

When feeding on Macrocystis it was noted that the animals eat only the

flattened strap-like portions of the thallus. This results in large piles of stipe

being left on the beach, along with the inflated basal portions of the fronds

(Fig. 19a). When feeding on Laminaria or Ecklonia it was noted that the animals

congregate on the flattened blades which are eaten from the edges (Figs 19, 20).

This preference for the flattened portions results in a typical grazing pattern, i.e.

many stipes denuded of fronds (Fig. 20a). Should food be scarce, however, the

animals can be seen feeding on the stipes of these two algae. An examination

of the mouthparts makes it obvious that feeding on a broad solid surface with

no edge on which to start is more difficult than feeding on flattened straps.

Animals feed apparently indiscriminately on both moist and dry algae. Feeding

432 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 19. a. T. granulatus feeding on blades of Macrocystis. b. T. granulatus concentrated around

the blades of Ecklonia.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 433

Fig. 20 a. Grazing pattern of 7. granulatus. Denuded stipes in foreground, ungrazed stipes in

background. b. 7. granulatus feeding on the edges of Laminaria blades.

434 ANNALS OF THE SOUTH AFRICAN MUSEUM

on dry algae, although seemingly difficult, can be explained. During the day the

kelp is tough and leathery, but at night with no heat for evaporation and with

moisture being deposited by dew and sea mists, the fronds soften slightly.

Although the juveniles are occasionally found feeding with the adults on

the larger brown algae, the majority are found on the debris lines. Their pre-

sence here is to some extent due to a behavioural trait. The juveniles are usually

found lower down on the beach than the adults, and often are caught in the

waves. Unlike the adults which flatten out to let the water flow over them, the

juveniles roll up into a ball, thereby capturing a bubble of air between the

pereiopods, making them buoyant. They are carried up the beach by the wave

and are left behind on the debris line along with their food. This latter in the

juveniles seems to consist mainly of animal matter. If on any one night the

number of juveniles feeding on different material is observed, the following sort

of numbers may be obtained: jellyfish fragments 17, U/va 5, other substances 5.

Should any fish or Physalia be washed ashore, larger numbers of juveniles are

found feeding on them.

FOOD PREFERENCES

To indicate very approximately the food preferences of 7. granulatus

(presumed to be similar to 7. capensis) the following procedure was followed:

Six containers measuring 400 x 200 mm and 100 mm deep were filled with

damp beach sand to a depth of 80 mm. Ten adult Ty/os were placed in each

container. Shallow containers were used to allow for a minimum of time to be

spent by the animals in burrowing, and for their easy removal. The containers

were kept under normal external light and temperature conditions, sheltered

from rain. The containers were left undisturbed for three days, during which

time no food was provided. Thereafter food of various types, having been

weighed, was placed randomly on the sand in the containers at 1730 hours. A

container without animals was used as a control. At 0800 hours the following

morning, those pieces of food which had been eaten on were reweighed, the

percentage water loss also calculated. The percentage water lost from the

remaining piece of food was added to the weight of the remaining piece of food.

This total was then subtracted from the original weight of the food. In the first

set of tests, the animals were taken from a beach almost devoid of food. The

animals were then starved for a further three days. When food was then pro-

vided on two consecutive days, those animals which emerged ate large amounts.

After this, less food was eaten at any one ‘sitting’; the food was then provided at

intervals of two to three days.

First choice

Roughly equal amounts (between 0,8 and 2,5 g) of the following six sub-

stances were provided: two brown algae — Macrocystis and Laminaria, one green

alga—Ulva, one red alga—Gigartina sp., cabbage, and fresh fish. All these

substances have been observed to have been eaten by Ty/os at the Blouberg

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 435

population studied. From Tables 5 to 9 it can be seen that the total amounts of

food eaten in descending order are: fish 19,032 g, Ulva 5,664 g, Laminaria 2,161 g,

cabbage 1,889 g, Macrocystis 1,481 g, Gigartina 0,298 g. If the number of times

a particular food was chosen out of the 30 times the food was provided is con-

sidered, the following emerges: fish 19, Ulva 16, cabbage 10, Laminaria 2,

Gigartina and Macrocystis one each. The three foods most preferred are thus

fish, Ulva, and cabbage, none of which appears regularly in the natural habitat.

The three which do appear frequently on the beaches were eaten only on the

first night, i.e. when the animals were starved.

It may be assumed that the location of food by Ty/os is partly by olfactory

means, as fish, U/va, and cabbage all have some odour, whereas the red and

brown algae are practically odourless when fresh. The second pair of antennae are

extremely well developed, and it has been demonstrated in Ligia (Hewitt 1907),

that these antennae have an olfactory function.

Second choice

As was frequently observed, Ty/os will feed on dry brown algae even when

fresh were available. This aspect was investigated by means of a preference test,

done in the same way as the previous test, but using only two containers of 10

animals each. In each was placed a piece of fresh and dry alga (Ecklonia) of

approximately equal weight. The dry alga was weighed at midday in all cases,

so that the water factor would be of minimal importance. The controls show that

if weighed at midday, any water taken up during the previous night was again

lost by the time of weighing. The dry controls remained at almost constant

weight. The percentage water loss for the fresh alga was not calculated, as the

fresh alga was never eaten. From Table 10, it can be seen that after six separate

feeding sessions, only the dry alga was eaten. An explanation for this preference

is probably that the mucilaginous exudation of the fresh alga is too sticky for the

mouthparts to cope with, and is thus avoided.

Food preferences of juvenile Tylos

As the juveniles are usually found on the debris lines, where the smaller

pieces of plant and animal matter wash up, it was decided to investigate the food

preferences of these forms. Two containers of 10 animals each (ranging from

8 to 20 mm) were used. The choice of food given was fish, Ulva, Macrocystis,

cabbage, and Plocamium (a small finely divided red alga). The results of this

test may be seen in Tables 11 to 14. From this it would seem that as in the adults,

the juveniles have a preference for animal matter (0,650 g fish eaten) but that

almost similar amounts of U/va and Plocamium were eaten (0,305 g and 0,376 g

resp.). Cabbage was seldom touched, and Macrocystis not at all.

STRUCTURE OF THE ALIMENTARY CANAL (See Fig. 22)

The two main references dealing with the alimentary canal of isopods are

Murlin (1902) and Nicholls (1930), both of which deal with forms such as

436 ANNALS OF THE SOUTH AFRICAN MUSEUM

BN Greist)

.% ‘* 150 ‘ v \d

Fig. 21b.

Fig. 21. Tylos granulatus. a. Cross-section of oesophagus.

b. Portion of wall of oesophagus; b.m.— basal membrane;

c.—cuticle; c.m.—circular muscle; ep.—epithelium; I.m.—

longitudinal muscle; s.—setae; s.g.—salivary gland

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 437

Ligia, Porcellio, Oniscus, and Philoscia. The only work dealing with Tylos is

that of Barnard (1925) which contains a superficial description of the ‘stomach’

of the animal.

In general terms 7Jy/os conforms with the overall isopod pattern, but has

several differences, which are mentioned below in the relevant sections. The

mouth is situated dorsal to the oral appendages, and just below the frontal

lamina of the head. The mouthparts consist of one pair of mandibles, two pairs

of maxillae, and a pair of maxillipeds. There is no obvious buccal cavity, the

mouth leading directly into the short oesophagus, which together with the

gastric mill constitutes the foregut. The walls of the oesophagus are strongly

convoluted and are covered with very fine backwardly directed setae. Externally

the oesophagus is surrounded by glandular material of the rather ill-defined

salivary glands. These latter open into the oesophagus near the mouth. In cross-

section (Fig. 21a, b) the oesophagus is roughly square, with the main masses of

longitudinal muscles concentrated at the centres of the sides of the square. The

chitinous cuticle has outgrowths in the form of numerous setae, and is secreted

by a single layer of epithelial cells just beneath it. The entire oesophagus is

surrounded by circular muscle fibres which tend to be arranged in large bundles.

Bands of muscles stretch from the oesophagus wall to the exoskeleton of the

head region, the salivary gland tissue being situated between these bands.

tergum

mid gut

Salivary gland

oesophagus

mandible

uropod hepatopancreas

P B2tolp

maxilliped

Fig. 22. Tylos granulatus. Lateral dissection, showing position of alimentary canal.

The oesophagus opens into the ‘stomach’ which contains a very efficient

gastric mill (Fig. 23), made up of the following structures:

1. The lateral ampullae. These are projecting structures, the median faces of

which meet above the opening of the oesophagus. These median faces have a

= =

438 ANNALS OF THE SOUTH AFRICAN MUSEUM

ventro-medial canal

oesophagus

bristle plate

Jateral ampulla

postero-ventral =

ampulla iss ee, Laren

medio - dorsal

ampulla

| a WM

antero- ventral

lamella

intestine

dorsal lamella ‘

hepato- pancreas opening

postero-ventral lamella

filter apparatus

ee ee A ee ae

4mm

Fig. 23. Tylos granulatus. Structure of the ‘stomach’.

trilobed chitinous structure possibly used for grinding. Ty/os lacks the small

antero-lateral ampullae found in Ligia and related isopods.

2. The median anterior ampulla, situated on the ascending anterior wall. This

is a lobed structure fringed with setae.

3. In the floor of the foregut, below the lateral ampullae, is a pair of crescentic

bristle plates, formed by the fusion of many bristles. The anterior portion of

these plates meet in the midline but diverge posteriorly. Posterior to these

plates are the ventral lamellae, while between them is the ventro-median

channel.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 439

4. The ventral lamellae may be divided into an anterior and a posterior region.

The anterior region, the margin of which is fringed with setae, bears on

each side a pair of plates, not mentioned in the description of Ligia or any

other isopod. These may be termed the postero-ventral ampullae. The latter

have a horizontal crescentic plate, as well as a medio-vertical plate, which

forms part of the wall of the ventro-medial. The posterior ventral lamellae

project into the midgut in the form of two rounded lobes armed with short

bristles.

5. In the ventro-medial region, just opposite the medio-vertical plates of the

postero-ventral ampullae, is an elongate spindle-shaped structure, the filter

apparatus. The anterior portion has a pair of lateral plates which work

against the medio-vertical plates of the postero-ventral ampullae. The

posterior portion projects between the lobes of the ventral lamellae, and is

equipped with numerous setae. This filter apparatus divides the ventro-

medial channel into two portions, each of which ends posteriorly at the

opening of the hepatopancreas.

6. Situated dorso-laterally and joined along the midline are the two lobes of

the dorsal lamina. These are acutely triangular in shape, consist of two layers

of chitin, and project into the midgut.

The hepatopancreas

As aiready mentioned, the gland on each side opens into the posterior por-

tion of the ventro-medial channel. Each gland consists of three tubular monili-

form portions, two postero-lateral and one antero-lateral (Fig. 22). The postero-

lateral portions stretch back along the mid- and hindgut, almost to the last

segment. The short antero-lateral portion stretches obliquely upwards on either

side of the ‘stomach’. In section (Fig. 24) the hepatopancreas can be seen to

consist of a single layer of glandular epithelial cells, around the central lumen

of the gland. In places the wall of the gland appears to be more than one layer

thick but this is merely due to the convolutions of the gland. The cells are

elongate with a highly granular cytoplasm and large deep-staining nuclei. In

the distal portion of many of the cells, large vesicles may be seen. These are

probably droplets of the secretion of the cells. In addition, many of the cells

have large clear central vacuoles. The outer surface of the hepatopancreas has

a very thin layer of connective tissue, between which and the glandular cells, very

thin circular muscle fibres may be seen. This muscle layer may be continuous

but is difficult to assess, being very thin.

The midgut

The gastric mill ends at about the level of the first thoracic segment and

opens into the capacious midgut, which stretches back to meet the hindgut at

the level of the tenth segment. There is no trace of a typhlosole, as found in

Ligia. In cross-section (Fig. 25) it can be seen that the midgut is lined with a

very fine chitinous intima, in which there is no sign of holes or ostia. Murlin

440 ANNALS OF THE SOUTH AFRICAN MUSEUM

a ig Was

¥ a

” “

+ $5

Fig. 24. Tylos granulatus. Cross-section of hepatopancreas; c.—connec-

tive tissue; n.—nucleus; va.— vacuole; ve. — vesicle

(1902) mentions the presence of very fine pores in the intima of the midgut

of several terrestrial isopods, including Oniscus and Asellota. Nicholls (1930)

could not find these holes in Ligia, neither were they present in Armidillidium,

Porcellio, or Idotea (McMurrich 1896). Beneath the intima is a single layer of

elongate epithelial cells. These are large cells, about 150 u in length, and have

large deep-staining nuclei. These epithelial cells are remarkable for their large

vacuoles surrounded by a thin granular cytoplasm. Numerous granules are visi-

ble in this cytoplasm, as well as minute droplets. These may be absorbed diges-

tive products. The outer bases of the epithelial cells are surrounded by a very

fine basal membrane. The outermost layer consists of a sheath of connective

tissue, with separate bundles of oblique muscle fibres.

The hindgut

The midgut opens into the hindgut at a constriction. The hindgut which

forms about one-quarter of the total gut length, occupies the last four segments

and opens at the anus, which is covered by the valve-like uropods. The inner-

most layer of the hindgut is a very fine chitinous intima, beneath which is found

a single layer of elongate epithelial cells, similar to those of the midgut (Fig. 26).

These cells contain large vacuoles, but not as obvious as those of the midgut.

The cytoplasm does not have many inclusions. The epithelial layer is very con-

voluted, with connective tissue between it and the next muscle layer. The muscu-

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 44]

ihe

gs ais

hel ~

“”

150 }*

Fig. 25. Tylos granulatus. Cross-section of midgut; b.m.— basal membrane;

c.i.—chitinous intima; g.— granules; n.—nucleus; 0o.m.—oblique muscles;

va. — vacuole

lar sheath surrounding the hindgut consists as circular as well as longitudinal

fibres. The circular fibres form a continuous band around the gut, while the

longitudinal fibres are arranged in separate bundles on the circular muscle band.

THE PROCESS OF DIGESTION

The progress of food through the gut involves several stages, some being

not particularly clear. The food is manipulated and eventually cut up by the

mandibles, then passes into the oesophagus in the form of fairly uniform pieces.

The posteriorly directed setae of the oral region all assist the food in moving

into the oesophagus, and prevent it from slipping out. The fine hairs of the

oesophagus have a similar function. While passing into the oesophagus the food

is mixed with the secretion of the salivary glands. (The exact nature of these

secretions is uncertain, as very refined techniques are needed to collect and test

them. A certain amount of mucilaginous fluid is secreted.) Contraction of the

muscles of the oesophagus walls passes the food on to the gastric mill. As none of

the structures of this complex is particularly hard, it would seem unlikely that any

trituration occurs. Rather it would seem that the lateral ampullae by muscular

action force the food against the bristle plates. Similarly the postero-ventral

ampullae would work against the filter apparatus, again squeezing the food.

Both these actions would result in food plus tiny pieces in suspension, to pass

—————-

442 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 26. Tylos granulatus. Cross-section of hindgut; c.i.—chitinous intima;

c.m.—circular muscles; I.m.—longitudinal muscles

into the ventro-medial channel. Contraction of the hepatopancreas would force

its contents into this channel, to be mixed with the food-fluid, and allow the

enzymes present in the secretion to act. Large food particles would be pre-

vented from entering the channel by the posteriorly directed setae and allow the

food to pass only backwards into the midgut.

The site of absorption has been the subject of much controversy. It has

been shown that in Nephrops (Vonk 1960), absorption can and does take place

in the midgut. Several decapod crustaceans, including Astacus (Jordan 1929),

Nephrops (Yonge 1924) and Atya (Van Weel 1955), have been shown to absorb

substances through the hepatopancreas. Nicholls (1930) showed that in Ligia

absorption took place through the midgut (even though lined with chitin) and

also through the hepatopancreas. It is likely that a similar set-up occurs in

Tylos. With contraction of the hepatopancreas the secretions are forced into the

ventro-medial channel, and mixed with the fluid part of the food. Any excess

secretion is forced into the upper portion of the gastric mill and mixed with the

solid food. With relaxation of the hepatopancreas, the mixture of foods in the

channel is drawn into the lumen of the gland. Dissolved food in the form of

monosaccharides and amino acids may then be absorbed by the glandular

epithelium. Oil droplets are probably also taken up, as is known to occur in

Atya (Van Weel 1955). Contraction of the muscles of the gastric mill then forces

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 443

the solid food particles and hepatopancreas fluid into the midgut where a longer

process of digestion takes place, and the released substances absorbed through

the chitinous intima of the gut (Vonk 1960). Food is probably kept in the midgut

for a period of up to 48 hours, depending on the type of food. Observation shows

that an adult Ty/os having fed on Ecklonia does not feed on the following night,

and probably will not feed on the second night either. The ‘storing’ of the food

probably accounts for the sporadic emergence of the adults. If the midgut is full

of digesting food, there would be no need to emerge and feed.

The secretion of the hepatopancreas

From a perusal of the food list, it will be seen that Ty/os is omnivorous, with a

bias towards algae. Omnivorous invertebrates almost invariably possess all three

major types of digestive enzymes, viz. carbohydrases, proteases, and lipases.

With this in mind, simple qualitative tests were performed to establish the

presence of these enzyme types in Ty/os.

A homogenate of hepatopancreas gland was prepared in the following way.

The glands of 40 animals were removed, weighed, and homogenized with a

phosphate buffer of pH 7,0. The homogenate was then centrifuged and the

supernatant further diluted with buffer to give a 10°, homogenate solution.

Carbohydrases: tests were performed for three carbohydrases, cellulase,

amylase, and sucrase. In the case of cellulase, a positive control in the form of

homogenate of Helix midgut was used, as well as a negative control in the form

of boiled hepatopancreas homogenate. No cellulolytic activity could be detected

in the active Tylos homogenate and it was concluded that no cellulase was pre-

sent. Amylase was tested for by means of starch solutions and Fehling’s test for

reducing sugar. The latter test was positive, indicating the presence of some

amylase-type enzyme. Sucrase was tested for, a positive control in the form of

invertase solution being used. A positive Fehling’s test indicated the presence

of a sucrase-type enzyme.

Lipase was tested for by means of the action of the homogenate on milk,

phenolphthalein and NaOH being used as indicators. Fatty acids were released

indicating the action of a lipase.

Protease was tested for by the use of congo red fibrin at pH 7,0. Only after

48 hours did any red coloration appear in the experimental tube, indicating that

some fibrin has been digested to release the red particles. The test was repeated

using a phosphate buffer of pH 8,0. After two hours the experimental tube

contents became red. In both tests, a dilute solution of trypsin was used for

comparative purposes. It was concluded that some protease was present which

operated in the alkaline range.

The results of the test for protease seemed to suggest that the pH of the

gut must vary if all the enzymes are to function close to their optimal pH.

Several animals were opened and the pH of the gut tested directly. The results of

this showed that whereas the pH of the hepatopancreas was 6,3, that of the

gastric mill and anterior part of the midgut is 6,9, while the posterior midgut has

444 ANNALS OF THE SOUTH AFRICAN MUSEUM

a pH of 7,4, and the hindgut 6,8. It would seem that the carbohydrases have

maximum activity in the foregut and anterior midgut, while the posterior midgut

with its higher pH would be the centre of protease activity. The acid pH of the

hindgut is probably due to the start of decomposition. The results are similar

to those obtained by Nicholls (1930) for Ligia oceanica. He found the pH of the

hepatopancreas to be in the region of 6,1-6,3, with the pH of the gut usually

higher than that of the gland. The pH of the intestine varied from 6,2 to 7,2,

while the rectum had an acid pH of 6,0.

RHYTHMICITY

In an attempt to gain more precise information regarding the rhythmic

activity of Ty/os, uninterrupted nightly observations were done on T. granulatus

for 36 consecutive days. Consecutive observations over a long period were not

carried out for T. capensis; interrupted observations over several weeks were

nevertheless made. So as to get an overall view of the situation, and to see trends

in the rhythm, all the information for T. granulatus was presented graphically,

together with other relevant factors. With reference to Figure 27, the following

should be noted.

The horizontal axis denotes the 24 hours of the solar day.

The vertical axis denotes consecutive days, the actual dates being given.

Solid black horizontal lines represent the time emerged by the animals.

Stippled areas denote dark of the sun, i.e. night.

Curved solid lines indicate low tide, the actual time of low tide being where

the line intersects the horizontal day line.

Curved broken lines represent high tide, the actual time of high tide being

where the line intersects the horizontal day line.

7. As the animals emerge and return over a period of up to 30 minutes, actual

time of emergence was taken when at least ten animals could be seen emer-

ging simultaneously. Similarly, time of return was taken when at least ten

animals could be seen returning.

Da 2

Discussion of Table 15 and Figure 27

From the above data several facts emerge, some of which can be explained

and others which can only be speculated about. The most obvious fact is that

T. granulatus is purely nocturnal and usually emerged at least one hour after

sunset. (T. Jatreillei by contrast is a diurnal animal, preferring damp shady

crevices of rocks. Pardi (1955) made use of this fact when investigating the

orientation mechanism of this animal.) The shortest time observed between

sunset and emergence was one hour. The shortest time observed between return

into the sand and sunrise was 24 hours. The mean time emerged for the 36 con-

secutive days of observation was 2,1 hours, the mode 2,0 hours. The longest time

out was about 23 hours while the shortest time out (apart from the exceptional

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE

hours

123 4 5 6 7 8 g 10 11 12 13 14 15 16 17 16 19 20 21 22 23 24

+ nm

died di Biedl,

ee | HWS

HWS

Fig. 27. Rhythmic activity of Tylos granulatus.

445

446 ANNALS OF THE SOUTH AFRICAN MUSEUM

13th and 25th) was 14 hours. The time the animals have to feed is thus very

short, and appears to be strictly controlled.

It can be seen that there is an overall trend to emerge later each night; this

gap between the time of emerging for successive days may be anything from

0 to 14 hours. Occasionally the animals emerge at about the same time for several

consecutive nights, e.g. from the 14th to the 18th inclusive. If the time for

re-entry for these days is noted, however, it can be seen that this becomes later

each day. If the time between re-entry is plotted against successive days, it can

be seen that a line may be drawn through these points at the 24,5 level. This is

the mean of the times between re-entry. The mode of this time gap, however, is

24,7 hours. Excepting the unusual days of the 13th and the 25th, the time between

re-entry on successive days is always more than 24 hours. If the cycle is one of

24,8 hours which is not unusual for littoral animals, there should be a correlation

between the time of emergence of the animals and the time of low tide (or the

time of high tide). With this in view, Spearman’s Rank Correlation Test was

done, using the three sets of consecutive 12-day observations and ignoring the

days in between. This test gave positive correlations of 1,0, 0,84, and 0,9. The

significance of this correlation was tested, using the ‘t’ test, which showed that

the probability of this correlation occurring by chance was less than 0,01, i.e.

there is a positive correlation between the time of low tide and the time of

emergence. It must be noted that although this test shows positive correlation

between tides and the cycles of activity, this does not imply a causative connec-

tion between these phenomena. It can be stated with considerable certainty that

T. granulatus possesses a 24,8 hour (lunar day) rhythm.

An explanation for the consecutive days on which the animals emerge at

the same time, may be that not all the same animals emerge, thus each day some

animals emerge later than others, and return to the sand later each day. Brown

(personal communication) states that in some sandy beach animals which

burrow, only part of the population emerges each day. Thus in the mollusc

Bullia digitalis it has been estimated that only 12°% of the population emerges

at any one tidal cycle.

If the times for re-entry and emergence are examined against the times of

high and low tide, it is seen that on five days out of the 36 the animals emerged

near the time of high tide, but only during neap tides. On all the other days they

emerge either at low tide, or during ebb tide. It is obviously advantageous to the

animals to emerge during a receding tide. If they emerged close to the time of

high tide, there exists the possibility of being swept to sea.

As has been shown, Ty/os has under natural conditions a lunar day rhythm.

It thus emerges a little later each night. If this system were invariable, eventually

it would not emerge until dawn. Being purely nocturnal, as already stated, this

would be most unsuitable. The animals correct for this effect of the lunar day

rhythm by a ‘switch back’ once a certain stage of the cycle is reached. This

switch back takes it to the early evening hours after sunset which then allows

many days of uninterrupted lunar rhythm. This switch back was seen to operate

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 447

on the 13th and on the 25th. On the former day a few animals were seen to

emerge at about 0300 hours, just on high tide. A vast number emerged at

2045 hours, whereas on the previous day the majority emerged at 0250 hours.

It was obviously important not to emerge at 0300 hours on the 13th, 1.e. at high

tide, as the tides were approaching springs. The animals would be low down on

the beach, and the increasing height of the high tide would certainly endanger

them. Emerging at 2045 hours brought them out at low tide when it was quite

safe for feeding.

Twelve days later, the animals again switched back. A few emerged at 0015

on the 25th, but a great number came out at 2030 hours. Again the tide was

approaching springs and although on the 23rd and 24th they had emerged at

high tide, the tide was getting higher each day, rendering their late emergence

more dangerous. To cause this switch back, there must be some rhythm imposed

on the lunar day rhythm. In all probability this is a semi-lunar or 14-day rhythm.

This would keep the animals in phase with the spring tides, the period between

successive springs being 14-15 days. The fact that the switch back of the 25th

occurred only 12 days after the previous one does not disprove this, but may

represent an abnormality in the rhythm due to external factors. It is interesting

to note that in both the observed switch backs, these occurred four days before

the spring tide. That the rhythm is a 15-day one may be inferred if Figure 27

is examined. Twelve days elapsed between the start of observations on 1 February,

and the switch back of the 13th. The HWS occurred on the Ist. Three or four

days before this the animals would have switched back, i.e. on 28 January.

Further evidence for this is that on 28 January the animals were observed to

emerge at 2100 hours at Lambert’s Bay, 288 km up the coast. (As there are only

three 15-day cycles available, statistical verification is not possible. With a

minimum of six cycles, 90 days of observations would need to be done.)

As it has been shown that Tylos possesses a 24,8 hour rhythm as well as

possibly a 15-day rhythm, it remains to investigate whether these are inherent

or whether they are triggered by external factors. With this in view, a rather

unrefined experiment was performed. A population of 25 animals of T. granula-

tus was placed in a perspex/glassfibre container, having an area of 250 cm* and

a depth of one metre. This container was placed in a completely dark room,

having a temperature of 17°C, and a range of about 3°C. This room was kept

in constant darkness for 18 days. The only light was that of a torch used for

brief intervals for observation. Food in the form of moist Macrocystis and

Ecklonia was regularly provided so as not to be a limiting factor. The sand used

at the start was damp, and at irregular intervals small amounts of sea water were

added to prevent drying out.

The animals were left for two days to acclimatize before observations were

started. The animals in the darkroom were thus separated from the effects of

the tides, temperature fluctuations, and illumination affecting the animals on

the beach, which were used as controls. The emergence of the animals from both

the beach and the darkroom was plotted against a time axis (Fig. 28). At the

448 ANNALS OF THE SOUTH AFRICAN MUSEUM

hours

13) 14. 15) 16) 17 1BV19) 200-21.522)23'24 1) 2-3 475) 6 7/88) 40 4 2

23 i . 24

24 | _ 25

LLL

25 aes 26

om |

26 — 27

Pod

27 ae 28

er]

28 a 29

29 ee | 1

ays azz ee

1 2

2 3

3 EE 4

WIE |

A | a, 5

5 ae 6

ZZ.

6 7

| C2

7 — 8

Fig. 28. Rhythmic activity of Tylos granulatus.

same time, ten animals were kept in a container of sand, away from the beach

but under normal light and temperature conditions. These latter could unfor-

tunately only be observed for six days. These results are also plotted on Figure

28. On the 29th, after it was suspected that not all the same animals emerge each

night, the darkroom animals were marked with numbers painted on their dorsal

surfaces so as to record in greater detail the exact animals which were emerging.

(Marking animals on the beach is not practicable; the number feeding on

0,2 m? of kelp was often found to be over 300. A very rough estimate of the

number of animals in the study area at Blouberg is between 3 000 and 4 000.)

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 449

The results of these observations may be seen in Table 16. From the latter and

from Figure 28, the following observations were made:

1. For the period of observation, the experimental animals emerged only

during night hours, even though they were subjected to constant darkness.

2. The same animals do not necessarily emerge every night, and on some nights

none emerge.

It was found that on 22 March, 29 days after the experimental animals

had been placed in the darkroom, nine emerged at 1700 hours, i.e. two hours

before sunset. The rhythm which kept the animals nocturnal appeared to be

breaking down. The container with the animals was then placed in the open

where they received normal light and temperature conditions. The following day

the animals emerged at about 2100 hours, almost two hours after sunset. They

had thus corrected the time of emergence, once normal light conditions were

restored. It is interesting to note that the Blouberg population emerged at 2015

hours on the same day.

Discussion of Table 17 (Tylos capensis)

Even from the limited data available for observations on JT. capensis,

several facts may be deduced. As in 7. granulatus the present species is strictly

nocturnal and usually emerges at least one hour after sunset. The time of

emergence is again usually close to that of low tide, as in T. granulatus, while

the actual time emerged is from 1} to 24 hours. Thus a lunar day cycle of 24,8

hours may be inferred for this species. As the animals always emerged at or

around the low tide of the dark hours, and as the observations cover a period of

almost 24 months, it may also be inferred that the switch back mechanism,

controlled by a 15-day cycle is present in this species.

REPRODUCTION AND GROWTH

Reproduction in Tylos is geared to the littoral and almost terrestrial mode

of life. Spermatozoa must be transferred to the female without the assistance of

water. For this, the male is provided with copulatory stylets. These latter are

adaptations of the endopods of the second pair of pleopods. The basal portion

of each stylet lies parallel with the posterior margin of the last pereion segment.

At the midline, each stylet is bent at a right angle posteriorly. In the adult male

of T. granulatus this portion can be 3-4 mm in length, and is adpressed to its

opposite member. Each is concave on the inner face, the two together forming

a channel for the passage of the spermatozoa. (The presence of the stylets, which

begin to develop in the very small males, is useful in distinguishing the sexes.)

The spermatozoa are produced in the testes, three on each side, situated in

segments three and four, dorso-lateral to the gut, and are passed into the stout

vasa deferentia. In these the needle-like spermatozoa are clustered into sperm-

atophores. The male genital orifice on pereion segment seven is just opposite

450 ANNALS OF THE SOUTH AFRICAN MUSEUM

the bend in the copulatory stylets, and the spermatophores pass along this

channel to be transferred to the female. The actual transfer was never witnessed,

and could have taken place at the surface or beneath the sand. Many of the

males dissected from almost all the localities were found to have spermatozoa

in the vasa deferentia. The ovaries during the breeding season are large sac-like

organs stretching between segments two and eight and are situated above the

hepatopancreas. A short oviduct leads ventrally to the female genital orifice

which is situated at the base of the fifth pereiopods. Oostegites which form the

brood pouch for the eggs are not present in all the females. These presumably

appear after a moult, when the female has ripening eggs in the ovaries. Ovigerous

females of T. granulatus were found during January and February from several

localities. The minimum length of an ovigerous female is 36 mm, the maximum

length found, 41 mm. Up to 1969, two ovigerous females of T. granulatus were

known, from the mouth of the Schuster’s River, collected in February 1937.

Numerous ovigerous females have now been collected. The eggs are 1,6-1,8 mm

in diameter. Barnard (1940), writing about the first two ovigerous females,

states that the oostegites remain flat and that the brood pushes the sternum

upwards. This was found to be true in all the later specimens. Barnard mentions

that this forcing up of the sternum results in a degeneration of the females’

internal organs, and he concludes ‘In fact it would seem improbable that the

mother could recover after the escape of the brood’ (1940: 438). With this in

mind, ovigerous females were injected with formalin on capture. When dissected,

the alimentary canal and heart were found to be normal, only the ovaries being

degenerate. It would seem probable that Barnard’s conclusions were the result

of imperfect preservation.

Of T. capensis, this is the first record of ovigerous females; several were

collected from Strandfontein from November to January. The minimum length

of an ovigerous female is 20 mm.

From the beginning of September (in 7. capensis) and the beginning of

October (in JT. granulatus) the ovaries start to enlarge and egg formation com-

mences, even before oostegites become apparent. By the end of January (T.

capensis) or the end of February (7. granulatus) ovigerous females are no longer

encountered. In both species it was noted that ovigerous females were only

found above the high tide zone, up to 20 metres in 7. capensis. This is probably

connected with incubation of the eggs, as the sand above the high tide mark

has a fairly constant and slightly higher temperature than the sand lower down

the beach which is influenced by the temperature of the incoming tide.

Measurements of many individuals of T. granulatus and T. capensis from

Blouberg and Strandfontein respectively, if plotted in the form of histograms

(Figs 29, 30), show that there are at least four moults and possibly more in the

life history of the animals. Figure 29 clearly shows four peaks of size, and if the

ecdysis for the very small individuals in the brood pouch is added it seems

probable that there are five moults, and that only in the final one do oostegites

appear in the females.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 451

7a fof

= 8

4 Soe 2a MiG 200) 24 2G 090) 136.0. ao. wan. ag

length mm

Fig. 29. Tylos granulatus. Histogram showing size grouping of males and females.

ASSOCIATION OF TYLOS WITH OTHER SPECIES

Tylos granulatus

As a nocturnal inhabitant of sandy beaches, this species would seem at first

to have very little contact with other animals. On closer examination a few

associations emerge, but one important factor remains unexplained. After more

than two years of observation, no predator has been recorded. In the sand itself,

the only possible predators could be golden moles (Chrysochloris capensis)

which occasionally burrow along the upper beaches. Birds such as gulls (Larus

spp.) and cormorants (Phalacocorax sp.) are active during the day when the

452 ANNALS OF THE SOUTH AFRICAN MUSEUM

number of animals

4 8 12 16 20 24 28 32 36 40 44 48

length mm

Fig. 30. Tylos capensis. Histogram showing size grouping of males and females.

isopods are up to 400 mm beneath the sand. Even probing in the sand is unlikely

to expose any animals. It is perhaps possible that genets (Genetta sp.) and pole-

cats (Ictonyx sp.) from the surrounding bush could prey on the animals at night.

This is, however, pure speculation, no proof having been found. Perhaps this

species does not have any predators at all.

COMPETITORS

Apart from the competition for space in the sand, and for food amongst

the individuals of the species itself, only two food competitors have been

observed. These are the amphipod Talorchestia quadrispinosa and the carabid

beetle Platychila pallida. Talorchestia is often seen in great numbers around

decomposing kelp, but even in great numbers these animals represent a rather

unimportant competitive element, when compared with the amount of algae

eaten by the isopods. Platychila is a well known member of the sandy beach

community, and has a distribution very similar to that of 7. granulatus. This

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 453

pale beetle is about 20 mm in length, and is also nocturnal. These insects were

often encountered about 150 mm below the surface of the sand when digging

for Ty/os. At night they were often seen scampering over the sand. Possessing

relatively large mandibles, they are predacious, and although never seen actually

eating Tylos, they were often seen feeding alongside the juvenile isopods on the

debris lines. They were seen to feed on Physalia and other animal matter, and

thus represent a fairly important competitive element low down on the beach

where only the juvenile Ty/os feed. Another carabid beetle often found in the

sand with Tylos is Acanthoscelis ruficornis. Also a predator, fewer of these were

found than of the previous beetle. Pachyphalaria capensis, a tenebrionid beetle,

is often found with 7y/os, scavenging mainly on algae, but does not occur in

sufficient numbers to constitute an important competitor.

COMMENSALS

Barnard (1932: 218) noted the following, when dealing with 7. granulatus:

*A minute oligochaete lives among the pleopods. It is 2,5-3,0 mm in length,

whitish, with 4 bundles of straight and apically simple spines in each segment

usually 3 spines in each bundle. It may be termed Enchytraeus tylidus n.sp.’

Many specimens of this oligochaete were found on animals from all the locali-

ties listed. Samples of 10 adults and 10 juveniles (less than 20 mm in length)

from Blouberg and Lambert’s Bay were examined. The results are given in

Table 18. From this it can be seen that almost all the adults carry these commen-

sals, and that the juveniles are ‘infected’ only after they have reached a length of

about 20 mm. The worm may breed beneath the pleopods as structures resem-

bling egg cases have been found. On further investigation it was found that since

Barnard’s mention of the worm, it has passed unnoticed and has never appeared

in the Zoological Record. The oligochaete has been found only on T. granulatus

and is probably an obligative commensal, confined to this single species.

Another commensal was a tiny mite which was sometimes found between

the pleopods, but more usually beneath the extensions of the fifth abdominal

segment and beneath the uropods in the region of the anus, where the exo-

skeleton is relatively thin. No visible damage due to the commensal could be

seen in any of the animals examined. From one to four mites may be present

per isopod; of 25 examined, 9 carried the mites. Dr G. C. Loots (personal

communication) states that the mite is an undescribed species of the genus

Veigaia. He also notes that the members of this family usually play a predatory

role in micro-habitats, especially in rich soil. This is the first record of a member

of the family having been found commensal on an invertebrate. Dr Loots

suggests that the mites may feed on the faeces of the isopod; this would account

for their concentration in the pleon region.

Tylos capensis

As with the former species, no predator has been found for this species.

454 ANNALS OF THE SOUTH AFRICAN MUSEUM

The beetles Acanthoscelis ruficornis, Platychila pallida and Pachyphalaria capensis

have all been found in association with T. capensis in False Bay.

COMMENSALS

Mites of the same species of Veigaia as were found on the former species

of Ty/os were collected from specimens from Strandfontein, Breé River, and

Knysna. 11% of the Knysna isopods, 8°% of the Breé River, and 5°% of the

Strandfontein animals carried the mites.

Also found on 7. capensis were the third larval stage of some rhabditiform

nematode, but only from the Strandfontein population. The minute worms were

found in numbers of up to 100 per individual, between the pleopods, and caused

no apparent damage. Hyman (1951) notes that many of these nematodes are

organic detritus feeders, sometimes epizootic on invertebrates.

COMPARISON OF SPECIES AND DISCUSSION

From the present study it is apparent that the two species are very similar

both structurally and with regard to their behaviour. The most obvious difference

is that of distribution, 7. granulatus being strictly west coast, T. capensis strictly

east coast, with no overlap. The difference in sea temperatures for the east and

west coasts undoubtedly has an effect on the temperature of the intertidal area.

The higher sea temperature of the east coast may account for the earlier breeding

season of T. capensis. There are also obvious differences in the size of the adults

of the species, ovigerous females of 7. granulatus being almost twice the length

of those of T. capensis. Structurally the difference between the species is small,

and lies in the degree of granulation of the exoskeleton, and the development of

the ventral processes of the fifth pleon segment. Other differences which have

been noted are in the choice of beach and in the depth to which the adults

burrow. 7. granulatus is found only on exposed beaches and burrows to a depth

of up to 400 mm, whereas T. capensis only goes down to about 300 mm, and may

be found on both exposed as well as sheltered beaches.

Judging from the numbers especially in T. granulatus, this genus is very

well adapted to its ecological environment. This latter—the sandy beach, poses

several problems to its inhabitants involving wave action, stability of the

substrate, food supply, predators, temperature/light/humidity fluctuations,

tidal rhythms. By being nocturnal and by burrowing into the sand, the animals

have simultaneously solved several of these problems. There are no predators

on the beach at night, whereas if the animals emerged during the day they would

be vulnerable (particularly the juveniles) to being preyed on by sea birds. In

spite of the relatively large surface-temperature and humidity fluctuations,

beneath the sand these factors remain relatively constant. Emergence at night

ensures that the temperature of the surface has dropped, thereby reducing the

danger of desiccation. This latter factor is also reduced as the relative humidity

increases at night. Burrowing also reduces the risk of exposure to wave action.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 455

The most important adaptation of Ty/os is its possession of lunar rhythms.

By manifesting a lunar-day rhythm closely correlated with the time of low tide,

and by being nocturnal, the animals emerge during the nocturnal low tide.

This gives protection from wave action. Food in the form of organic debris is

washed ashore with the high tide and left stranded at the ebb. The animals can

thus feed quite safely. By restricting the time of feeding to two hours, a limit

is placed on the amount capable of being eaten. This is compensated for in

several ways. As the animals burrow at or near the high tide mark, the time spent

in searching for food is reduced. Being omnivorous and possessing all the major

types of digestive enzymes, there is no great need to be very selective with regard

to the type of food. The mouthparts are very efficient and the gut capacious,

thus a maximum amount of food may be eaten in the time available.

The possession of a semi-lunar rhythm ensures that the animals are kept in

phase with the spring tides. In this way animals may emerge on a rising tide

during neaps, but never during springs. Once again the danger of being swept

to sea is eliminated. To what extent the external environmental factors influence

the rhythms, and to what degree they are endogenous, is difficult to assess.

That the rhythms are to some extent endogenous, is demonstrated by the fact

that they continue to function under constant darkness, away from the sea.

Temperature would appear to have little if any influence on the rhythm. This

is necessary, for, to quote Brown (1959: 1539): ‘To possess adaptive signifi-

cance for the organism, a considerable degree of temperature independence is

needed to permit the cycles to retain their normal lengths when exposed to the

natural temperature fluctuations.’ It has been shown that Uca and Cambarus

retain their rhythm under constant temperatures ranging from 6 to 26°C.

Temperatures within two or three degrees of freezing point, however, are capable

of inhibiting the rhythm, through a slowing down of the metabolic rate. Brown

(1961) introduces another factor which has bearing on the maintenance of

rhythm. In Uca, animals which are kept together in the laboratory under

constant conditions exhibit rhythms with precision, but if kept separately

in individual containers, they show a gradually decreasing amplitude in their

rhythms. Once a separated individual is joined by another, the rhythm is restored.

This social factor has not been further investigated, and remains largely unex-

plained. It is likely that this has some bearing on the problem of Ty/os where,

on the beach under natural conditions, the rhythms are fairly precise and are

exhibited by the population as a whole rather than by the individuals, but in the

laboratory where only a small portion of the population is studied, the rhythms

are subject to greater variation. External light must have some controlling and

supplementing effect on the rhythms involved. Thus if low tide is close to sunset,

the animals would come to the surface, and only if the light intensity is low

enough will they emerge. This testing of the light intensity is probably the

function of the pause observed in all emerging animals. In this connection it

was noted that the animals seldom emerged at the time of full moon.

The possibility also exists that the rhythms are not endogenous, and that

456 ANNALS OF THE SOUTH AFRICAN MUSEUM

under so-called ‘constant’ conditions, animals such as Uca are still responding

to some less obvious external factor. It would seem from preliminary work

carried out by Brown and his co-workers that this external factor may be corre-

lated with barometric pressure, but is not this pressure itself.

SUMMARY

1. An introductory ecological/biological study of Ty/os granulatus and Tylos

capensis has been carried out.

2. Complete descriptions and synonymies for both species have been given.

3. The geographical distribution of 7. granulatus was established as being

from Cape Point to northern South West Africa, while that of T. capensis

was from False Bay to southern Natal.

4. It was found that both species inhabit sandy beaches of varying slope, and

sand of varying grain size.

5. Both species burrow in the sand around the high tide mark, moving up or

down the beach as the height of the high tide moves.

6. J. granulatus was found to burrow to a depth of about 400 mm, T. capensis

to a depth of about 300 mm.

7. Both species execute a helix-shaped path through the sand.

8. Entry point into the sand is marked by a low mound of sand, exit by a

shallow hole.

9. Both species were found to be omnivorous, with a leaning towards an

algal diet in the adult stage.

10. Food preference tests indicated that the adults and juveniles preferred

animal matter to vegetable matter, and that the adults preferred dry algae

to fresh.

11. The structure of the alimentary canal and associated structures were investi-

gated, as well as the process of digestion.

12. Both species were found to be strictly nocturnal.

13. The rhythmic activity of both species was investigated. Both were found to

possess a lunar-day rhythm, bringing them out of the sand at or close to

low tide.

14. A semi-lunar rhythm is postulated. This would account for a switch back

observed in the times of emergence. This mechanism militates against the

animals’ emerging at the time of high water of spring tide.

15. Oviposition and release of young was found to take place in midsummer,

i.e. December to February.

16. Five moults are thought to occur during the animals’ life span.

17. No predators were found for either species.

18. No important food competitors were found.

19. Both species were found to carry up to five specimens of a commensal mite

per isopod.

20. T. granulatus adults were often found to have several specimens of a com-

mensal oligochaete amongst the pleopods.

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 457

21. T. capensis from Strandfontein was found to carry rhabditiform nematode

larvae amongst the pleopods.

ACKNOWLEDGEMENTS

My sincere thanks are due to Professor A. C. Brown of the Zoology Depart-

ment of the University of Cape Town, under whom this investigation was

initiated, and to Dr D. van Z. Engelbrecht of the Zoology Department of the

University of Stellenbosch, who supervised the completion of this project, and

who gave useful advice and criticism.

My thanks are also due to the following friends and advisors, without

whose help, this investigation would have been the poorer: Mr C. G. Coetzee,

Director of the State Museum, Windhoek, for assistance with collecting along

the South West African Coast; Dr A. J. Hesse of the South African Museum for

identification of the Coleoptera; Dr G. C. Loots of Potchefstroom University

for identification of the mites; Dr N. F. Paterson of the South African Museum

for many helpful comments; Mr and Dr M. J. Penrith for their assistance with

collecting and observations; Mr R. Simons, Government Algologist, for assist-

ance with identification of the algae.

REEERENCES

ARCANGELI, A. 1953. Precisazioni ecologiche sui generi Ligia Fabr. e Tylos Latr. (Crostacei.

Isopodi terrestri). Atti Accad. Sci., Torino 87: 200-212.

BARNARD, K. H. 1924. Contributions to a knowledge of the fauna of South West Africa. III.

Crustacea, Isopoda Terrestria. Ann. S. Afr. Mus. 20: 231-236.

BARNARD, K. H. 1925. The digestive canal of isopod crustaceans. Trans R. Soc. S. Afr. 12:

27-36.

BARNARD, K. H. 1932. Contributions to the crustacean fauna of South Africa. 11. Terrestrial

Isopoda. Ann. S. Afr. Mus. 30: 179-388.

BARNARD, K. H. 1940. Contributions to the crustacean fauna of South Africa. XII. Further

additions to the Tanaidacea, Isopoda, and Amphipoda, together with keys for the identifi-

cation of the hitherto recorded marine and fresh-water species. Ann. S. Afr. Mus. 32:

381-543.

Brown, A. C. 1959. The ecology of South African estuaries. Part IX: Notes on the estuary

of the Orange River. Trans R. Soc. S. Afr. 35: 463-473.

Brown, F. A. 1961. Physiological rhythms. Jn: WATERMAN, T. H. ed. The physiology of Crus-

tacea. 2: 401-430. New York, London: Academic Press.

Buppe-Lunp, G. 1885. Crustacea Isopoda terrestria, per familias et genera et species. Hauniae:

Nielsen & Lydiche.

BupbpeE-LuND, G. 1906. Die Landisopoden. Dt. Siidpol.-Exped. 9: 69-92.

BuppDE-Lunp, G. 1909. Isopoda I: Land-Isopoden. Jn: SCHULTZE, L. Zoologische und anthro-

pologische Ergebnisse einer Forschungsreise in westlichen und zentralen Siidafrika. 2:

53-70. Fischer: Jena. Denkschr. med.-naturw. Ges. Jena 14: 53-70.

COLLINGE, W. E. 1945. Note on some South African terrestrial Isopoda. Ann. Mag. nat. Hist.

(11) 12: 344-347.

Do ttrus, A. 1895. Voyage de M. E. Simon dans |’ Afrique australe. Crustacés isopodes terrestres.

Mem. Soc. zool. Fr. 8: 345-352.

Haucuton, S. H. 1931. The late Tertiary and recent deposits of the west coast of South

Africa. Trans. geol. Soc. S. Afr. 34: 19-58.

458 ANNALS OF THE SOUTH AFRICAN MUSEUM

HewilrtT, C. G. 1907. Ligia. L.M.B.C. Mem. typ. Br. mar. Pl. Anim. 16: 1-37.

Hyman, L. H. 1951. The Invertebrates 3. Acanthocephala, Aschelminthes, and Entoprocta.

New York, Toronto, London: McGraw-Hill.

JACKSON, H. G. 1928. The morphology of the isopod head. Part II. The terrestrial isopods.

Proc. zool. Soc. Lond. 1928: 561-595.

*JORDAN, H. J. 1929. Vergleichend Physiologisches tiber Resorption. In: Handbuch der normalen

und pathologischen Physiologie. 4: 167-180. Berlin: Springer.

Krauss, F. 1843. Die siidafrikanischen Crustacea. Stuttgart: Schweizerbart.

*McMourricu, J. P. 1896. The epithelium of the so-called midgut of the terrestrial isopods.

L. Morph. 14: 83-108.

MATZAKIS, J. 1956. Observations éthologiques sur les Ty/os (Isopodes Onicoides) du Rousillon.

Vie Milieu 7: 107-109.

MEaD, F. 1968. Observations sur l’ecologie de Tylos latreillei Audouin (Isopode Tylidae) et sur

son comportement en milieu natural. Vie Milieu 19: 345-362.

MENzIES, R. J. 1952. The occurrence of a terrestrial isopod in plankton. Ecology 33: 303.

MuRLIN, J. R. 1902. Absorption and secretion in the digestive system of land isopods. Proc.

Acad. nat. Sci. Philad. 54: 284-359.

NICHOLLS, A. G. 1930. Studies on Ligia oceanica. II. The processes of feeding, digestion and

absorption, with a description of the foregut. J. mar. biol. Ass. U.K. 17: 675-705.

PANNING, A. 1924. Isopoda. Beitr. Kennt. Land.-u. Siisswasserfauna Dt.-SiidwAfr.2: 171-201.

ParpI, L. 1955. L’orientamento diuno di Tylos latreillei (Aud. & Sav.). Boll. Ist. Mus. Univ.

Torino 4: 167-196.

Soika, A. G. 1954. Ecologia, sistematica, biogeographica ed evoluzione del Tylos Jatreillei

Aud. (Isop. Tylidae). Boll. Mus. civ. Stor. nat. Venezia ’7: 63-84.

STEBBING, T. R. R. 1910. Report on the marine biology of the Sudanese Red Sea. 14. The

Crustacea, Isopoda, and Tanaidacea. J. Linn. Soc. (Zool.) 31: 215-230.

VANDEL, A. 1943. Essai sur l’origine, l’évolution et la classification des Oniscoidea (isopodes

terrestres). Bull. biol. Fr. Belg. suppl. 30: 1-136.

VANDEL, A. 1945. La répartition géographique des Oniscoidea (crustacés isopodes terrestres).

Bull. biol. Fr. Belg. 79: 221-272.

VANDEL, A. 1952. Le genre ‘Porcellio’ (crustacés, isopodes, Oniscoidea) évolution et systéma-

tique. Mém. Mus. natn. Hist. nat., Paris (n.s. zool.) 3: 81-192.

VANDEL, A. 1960. Isopodes terrestres. Faune Fr. 64: 1-416.

*VAN WEEL, P. B. 1955. Processes of secretion, restitution and resorption in gland of midgut

of Atya spinipes Newport. Physiol. Zool. 28: 40-54.

Vonk, H. J. 1960. Digestion and metabolism. Jn: WATERMAN, T. H., ed. Physiology of Crustacea

1: 291-316. New York, London: Academic Press.

*Yonge, C. M. 1924. Studies on the comparative physiology of digestion. 11. The mechanism

of feeding, digestion and assimilation in Nephrops norvegicus. Br. J. exp. Biol. 1: 343-389.

(* — original not seen)

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE

TABLE 1

Collecting localities for Tylos granulatus

B.K. = present author

Locality

Ventura wreck site, S.W.A.

Mowe Point, S.W.A.

Anichab, S.W.A.

Luderitzbucht, S.W.A.

Orange River mouth

Hondeklip Bay

Groenriviermond

Lambert’s Bay

Paternoster

Saldanha Bay, Langebaan

Ysterfontein

Melkboschstrand

Blouberg, Milnerton

Hout Bay

Noordhoek

Olifantsbosch

Witsands Bay

U.C.T. = University of Cape Town

Collector

B.K.

Panning, B.K.

Budde-Lund, Panning, U.C.T., B.K.

Ure EK

Batnard, UCP:

Barnard, B.K:, U.C.T.

Bake

Barnard, B.K.

TABLE 2

Collecting localities for Tylos capensis

Locality

Simonstown

Muizenberg

Strandfontein

Macassar Beach

Somerset Strand

Gordon’s Bay

Pringle Bay

Kleinmond

Wilderness

Breede River estuary

Keurbooms River mouth

Knysna

Plettenberg Bay

Whitney

Port Elizabeth

East London

Natal

Collector

Budde-Lund

Barnard, B.K.

B.K.

Barnard, Stebbing

Stebbing

B.K.

Barnard

B.K.

Barnard

BK UG

Barnard, B.K.

S.A. Museum

B.K.

S.A. Museum

Natal Museum

459

460 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 3

Weight of food eaten by adult Ty/los at single feeding session

Prefeeding | Postfeeding Weight Wt eaten oa

weight (g) weight (g) eaten (g) Body wt

12 15,26

jfa3) 20,71

2 22,83

0,9 16,47

1,0 24,73

ili 19,35

0,8 21,70

1,0 as)

0,9 21,82

0,6 17,95

TABLE 4

Food list of Tylos granulatus

(Only substances observed to be eaten included)

Vegetable matter Animal matter

Phaeophyta Haarder

Ecklonia maxima Cabbage Joseph shark

Laminaria pallida Onion Nudibranchs

Macrocystis pyrifera Orange Gastropod egg cases

Watermelon Horny branchiopod shells

Chlorophyta Bread Jellyfish

Bryopsis sp. Portuguese Man-of-War

Codium fragilae capense Red bait tests

Enteromorpha sp. Cormorant

Ulva sp. Tern

Rhodophyta

Aristothamnion purpuriferum

Botryoglossum platycarpus

Carrodaria virgata

Carpoblepharis flaccida

Ceramium planum

Champia lumbricalis

Gigartina stiriata

Gymnogongrus dilatatus

Gymnogongrus vermicularis

Phyllemenia hieroglyphica

Plocamium cornutum

Porphyra capensis

Pterosiphonia cloiophylla

Suhria vittata

Food list of Ty/os capensis

Phaeophyta Sour fig leaves Cormorant

Ecklonia maxima Ruppia maritima Haarder

Laminaria sp. Watermelon Hydroid perisare

Sargassum heterophyllum

Chlorophyta

Caulerpa sp.

Codium fragilae capense

Ulva sp.

Rhodophyta

Champia compressa

Pterosiphonia cloiophylla

Suhria vittata

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE

SSO] ]O1.U0d

Injen %

SSO'TI

Z0S‘0

1110

8670

191°Z

Igr'l

[210

Lacie O7S‘Z 956 I 0771

0000 000°0 000°0 088°0

beGia 07S ‘7 9661 O01'7

6770 991°0 1100

06 I v6 I vL6 I

= 6S1°C OOI 7 686 I

060°0

6£0'I

ae 6CI'T oe ce

srl

000°0

ey mg 18r'l =

¢ Vv € C

Snojnuvas Sojk], JO SoUdIZJo1d poo}

¢ alaV

Ysty

PATI

oseqqey)

BUIICSIH

BIICUIWIeT

SIJSADOIOVJI

xog

89/L/TE

ANNALS OF THE SOUTH AFRICAN MUSEUM

462

SSO] JO1]U0D

INeM %

86'S

OIS‘0

[2}0.L

O7S‘T vir‘ 9Er0 L680

6r8‘I 07S ‘Z C80°7 OOI*Z

p91 8070

CO00‘T oor‘ T

691‘T = = 899‘T

€50°0 OLI‘O

L8S‘0 8Z0'T

0r9‘0 == a S6I‘T

G v € Z

SnjonuvdAS SO]AT, JO SadUdIOJoI1d Poo

9 alav

Orl'T

€6L‘0

€£6'1

OI1‘O

619°T

6CL'T

860°0

L6L‘0

6680

ysi4

PAIN

oseqqery

BUI}ISIDH

elLIVUIWIe T

SIJSADOIOVIY

xog

89/8/T

fee)

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE

SSO]

oyem %

L60°0

9691

€6L‘I

6910

190°1

OFT I

£500

jo1ju0D

906°0

ETL‘O

v7T0 ILZ‘0

96L‘I L09‘T

070° = 8181

¢ v €

SnIDINUDAS SOjA], JO SAdUdIOJOId POO{

L aiavy,

Ysty

BAT

oseqqey

BUIJIVSID

RlIIVUIWIe T

SIJSADOIOV IAI

xog

89/8/7

any

eal

en)

by,

eo)

464

6€ CLG

eT 76r' I

SSO] {OlU0D

INVM %

rIl‘o

S8I0

SJO]NUDAS SOJAT, JO S9dUdIOJOId Poo;

8 alaVv

ysiy

PAT

oseqqey

BUILIVZIDH

eIIVUIWIE T

SI}SADOIOR IY

xog

89/8/¢

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE

SSO]

IyeM %

jo1Uu0d

610°0

bre'y

SOr'T

Lv8‘0

TST TC

99L‘0

—

Gun

8810

006°!

610°0

16L'I

O1s'T

0070

——

VEEL

bes’ I

snyojnuvas soja, JO SsoUdIEJo1d POO}

6 ATEVL

ystd

PATIL

oseqqry

VUIJIVSIO

elICUIWIeT

SIysADO IORI

xog

89/8/8

466 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 10

Dry/fresh brown alga (Ecklonia) preference test

Date Box i “Box 2 Control Total eaten (g)

16/9/68 Dry 1,413 1,455 1,754

1,263 1,268 —

0,150 0,187 —

Fresh 2,445 2,647 2,611

18/9/68 Dry 1,475 iL 357/7 1,580

1,398 — —

0,077 — —

Fresh 2,424 1,956 2,192

20/9/68 Dry S860 S161 1,863

1,381 1,453 —

0,205 0,158 —

Fresh 1,932 2,289 1,849

22/9/68 Dry eat 1,623 1,695

— 1,358 —

— 0,265 “=

Fresh 1,981 2,035 2,107

25/9/68 Dry 1,349 1,392 5l2

1,219 —

0,130 —- —-

Fresh 2,193 2,401 2,385

28/9/68 Dry 1,604 1,449 1,449

— 1,385

0,064

Fresh 2,035 1,991 1,941

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 467

TABLE 11

Food preferences of juvenile Tylos

Control Total % Water

15/9/68 Box 1 Box 2 (g) eaten (g) loss (g)

Fish 0,539 0,616 0,602

0,495 0,488 0,599

0,044 0,128 0,003 0,172 4,2

Ulva 0,441 0,429 0,460

0,368 — 0,419

0,073 0,041 0,073 9,1

Plocamium 0,504 0,481 0,490

0,464 0,449 0,464

0,040 0,032 0,026 0,072 5,4

Macrocystis 0,771 0,754 0,798

Cabbage 0,681 0,703 0,665

— 0,642 0,611

0,061 0,053 0,061 8,0

TABLE 12

Food preferences of juvenile Tylos

Control Total % Water

17/9/68 Box 1 Box 2 (g) eaten (g) loss (g)

Ulva 0,388 0,413 0,409

0/327) 105363) 0;367

0,061 0,050 0,042 0,111 10,3

Plocamium 0,453 0,420 £0,484

— 0,381 0,463

0,039 0,021 0,039 4,4

Macrocystis 0,683 0,703 0,698

Cabbage 0535.) (05613, 71105550

Fish 0,477 0,522 0,492

0,369 0,496 0,474

0,108 0,026 0,018 0,134 3,8

468

ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 13

Food preferences of Juvenile Tylos

Control Total % Water

19/9/68 Box1l Box2 (g) eaten (g) loss (g)

Ulva 0:325 05369" 0410

0,309 0,388

0,060 0,032 0,060 8,0

Plocamium 0,503 0,486 0,471

0,420 0,445 0,449

0,083 0,041 0,022 0,124 4,8

Macrocystis 0,705 0,720 0,684

Cabbage 0,489 0,532 0,516

0,471 — 0,477

0,018 0,039 0,018 TS

Fish 0,613 0,586 0,590

0,526 0,484 0,570

0,087 0,102 0,020 0,189 3,4

TABLE 14

Food preferences of juvenile Tylos

Control Total % Water

22/9/68 Box 1 Box 2 (g) eaten (g) loss (g)

Ulva 0,440 0,419 0,398

0,397 0,401 0,360

0,043 0,018 0,038

Plocamium 0,621 0,570 0,597

— 0,529 0,574

0,041 0,023

Macrocystis 0,580 0,694 0,689

Cabbage 0,388 0,412 0,370

— 0,389 0,340

0,023 0,030

Fish 0,664 0,580 0,683

0,094 0,061 0,031

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE 469

TABLE 15

Data for 36 consecutive days observation of Tylos granulatus

Hours

Date Time Time Hours between

emerged returned out returns

1/2/68 2050 2230 2,50

D 2100 2300 2,00 24,67

3 2LTS 2330 2,50 24,50

4 2210 0015 225 24,83

5 2230 0030 2,00 24,16

6 2315 0130 225 24,83

8 0005 0210 2,00 24,67

9 0040 0230 2,00 24,33

10 0125 0305 1,50 24,58

11 0200 0330 1,50 24,41

12 0245 0420 1,50 24,83

13 0300 0330 0,50 23,16

2045 2220 1,50 18,50

14 2045 2215 1,50 23,91

15 2045 2230 2,00 24,25

16 2045 2240 2,00 24,16

17 2045 2240 2,00 24,00

18 2045 2250 2,00 24,16

19 2030 2310 2,50 24,33

20 2150 2350 2,00 24,67

21 2200 0015 DDS 24,41

Dip 2245 0110 2,50 24,91

23 2340 0210 DDS 25,00

25 0020 0040 0,34 22.50

2030 2230 2,00 21,83

26 2040 2310 2,50 24,67

Di 2100 2240 75 23,50

28 2120 DSS 2,00 24,58

29 2120 2345 DOS 24,50

1/3/68 2130 0020 2,70 24,58

2 2230 0100 2,50 24,67

3 2300 0130 2,50 24,50

5 0010 0210 2,00 24,67

6 0015 0215 2,00 24,08

if 0030 0220 2,00 24,08

8 0045 0250 2,00 24,50

470

TABLE 16

ANNALS OF THE SOUTH AFRICAN MUSEUM

Data for observations on darkroom animals of 7. granulatus

Date

24/2/68

1/3/68

COND

2120-2145

2115-2215

0030

0005

2250-2330

0045-1000

2040-2200

0100

0250

0105

Time

emerged

Time emerged | Time returned

0030

2320

0150-2000

0040-0115

0156-0210

2230-2330

0315.

0125-0130

TABLE 17

Time

returned

Individuals

8 animals

3 animals

13:2.-3.960.77, delegioe

16, 18, 21, 20

2.3. 6, 8,9 es os

1, 132 1S ats

4, 6,-8, 10; Be ts:

15518520

24, 5,i/6, 7,105 tae

12, 13., 14S ae

135719520

2535 D5 Os TeaObees

10; 11; 12, 1504s

fS, 17; 18, 2082

Data for observations on Tylos capensis

15/1/70

16/1/70

19/1/70

25/1/70

28/1/70

31/1/70

4/2/70

9/2/70

11/2/70

19/2/70

27/2/70

5/3/70

6/3/70

9/3/70

BIOLOGY AND ECOLOGY OF THE GENUS TYLOS LATREILLE

TABLE 18

Numbers of oligochaete Enchytraeus tylidus present on adult and

juvenile Ty/os granulatus

Locality Number of oligochaetes per individual Ty/os

EE aL

Blouberg No. adults — 2 5 — | | ]

No. juveniles 9 1 — a -— —- —

Lambert’s

Bay No. adults — 2 2 3 2 1 —-

No. juveniles 7 2 — 1 -—— | — —

47]

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below.

Figure captions and tables to be on separate sheets.

ILLUSTRATIONS

To be reducible to 12 cm x 18 cm (19 cm including caption). A metric scale to appear

with all photographs.

All illustrations to be termed figures (plates are not printed; half-tones will appear in their

proper place in the text), with arabic numbering; items of composite figures to be designated

by capital letters (A, B, C etc.).

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BuLLouGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. J. Conch., Paris

88: 100-140.

FISCHER, P.-H., DuvAL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des

littorines. Archs Zool. exp. gén. 74: 627-634.

Konn, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 19606. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn: SCHULTZE, L.

Zoologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und

zentralen Siid-Afrika. 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16:

269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature

issued by the International Trust for Zoological Nomenclature (particularly articles 22 and

51). The Harvard system of reference to be used in the synonymy lists, with the full

references incorporated in the list at the end of the article, and not given in contracted form

in the synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

Brian Kensley

ASPECTS OF THE BIOLOGY AND ECOLOGY OF

THE GENUS TYLOS LATREILLE

Pe —_ wr ie . al

pane 1 )

é ; .

}

i" \

ks i oe

4 in

‘

| .

ie»

i i

a =

ey J

ak J

’ op

‘ 5 ; d

7 ry

: 4 or dae

=] ay

= a: 1

P a at

4 ti y

E : f

j a

= ri ‘

a ‘ é

\ r

Q J

ey UP

‘

au 2

{ 1

E nM

= aa

i ¢ “+ :

| See ee

YTUTION NOILNLILSNI NVINOSHLIWS

TEE ATES GO OMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLINS S31uvugiE

S. *. : i |

NVINOSHLIWS

SMITHSONIAN

NVINOSHLIWS

SMITHSONIAN

NVINOSHLIWS

S3!IYVYGIT LIBRARIES SMITHSONIAN INSTITUTIO®

f 2 i 2 ” | =

Lu Ww ,

- — a = a Wi _

< x ba = < Sof

oO anf ~ ne few 7

co Oo a) — a. oe

call = — Fr oy 2

YVudii LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3ZIYNVWY q

c = lp = s z

70 ee Ps) 9 aa i

=) 3 aa

> kK 5 > kK > en

= - = F .

> zs A Z ‘e a

ITUTION NOILALILSNI NVINOSHLINS S3ZIYVWHRIT LIBRARIES) SMITHSONIAN

nt Z 2 z & eo Ov oaee =

> = op z= 4 ae % rar >

\, 0 G4 r oy 1 AA O +

‘es Vi E 2 E x 2 E

> KG S > = > =

x FA 4 w” Pua (ep) * Prk Tp) 2

HVYGIT_ LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS Saluvugl

a uw o i So Ss #

i % pat S es.

= = c n a A <<

= mo = m = Sm

S) si Zs S) = rs) =

= al a aoe Pa Roe i

NOILALILSNI NVINOSHLIWS S41uvu git LIBRARIES_ SMITHSONIAN _INSTITUTIC

za. \ | ficaas : :

O cm fo) = O 4 sa

s x Ya E : iy:

= > ASE > © il

rai Cay * \ ra “4 rae bh ny

”) m S\S he m als m

= on pe = w = w

yYWugiq LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S31yu¥vVug

. a * W” Bo mie ” z n .

z =| z XK ie z =

js : 5 NN G a 5

z= 2 =z WW sg. = 2

= > ee! ee = >

”) = “” Ps ” =

NVINOSHLINS SJ yVY Gil LIBRARIES SMITHSONIAN INSTITUTIG

ee = w” = a) =

vf a Ww ra ul a

Be ei fag 2 o a

< = eee C =< ae

me 7 re 4 a ee

tL) iS) vd O = fe)

| uy = wal Zz sal Z

NYVYUAIT VI BRARIES _ SMITHSONIAN INSTITUTION _NOILNLILSNI NVINOSHLINS S3IYVWY gq

; rT | faa ; os q

_ oO eh Oo ~ ae oO THSVAT

to ee: a 8) noe w +k pea vo

20 5 ce = aap \. ROS

> 'y > ke > bk

79 Es a ee - =

h Z S Z ; i

ITUTION NOILALILSNI NVINOSHLINS S31IYVYRIT LIBRARIES SMITHSONIAN

" wn = wn z Sas 2) Fa

NS x es = Rae = =

NS i = ” x S

NaS oe a” Dy n D 7)

| = = > = SO. PS

a ar tan ee Peene ee ey pe

7 ‘SMITHSONIAN INSTITUTION NOILMLILSNI_NVINOSHLINS S31uvugIT LIBRARIES. St

wn Ph oF ‘e wn”

= Site z < z ae =

. = x =| = 4 = +4 .

a 3 se \ai $2: s YF . § ZT

WKH oO ‘ } A " w bp .

S : E Wz. : oY A ORNS

» 2 = ahs. >" = >" . = ~\* >

e447 7) its Zz 7) 2 an a Zz

eee OS, Sat SVE8 I LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_N

i z G " Wy? “ a

ar ac pe oc a. oa ae

Cc c <x ii | < Cc

- & e 5 co - = co O

t a om 2 % S ur a

RIES INSTITUTION. NOILNLILSNI NVINOSHLINS S3ZIYVYGIT_LIBRARIES S

= oo = w . o oo =

Gar: Lf , a = P) ’ > 2 wm =

42) = a ath ‘: _ 4

Ys) -§ Wy = = = ie a

ty» Ww 3 —

Pi @ ; Z ? i: zZ |

JLSNI NVINOSHLINS S3!1YVYEGIT LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI_ NN}

- 8 & 2 et) eas ae ame 2 =

S r if. Vy 5 = re N z til ‘* =

. 8 Gdg % g : Ef ff

. & <3 “yy = z = 2 Ty E

1} = > ; = i ae = > | =

a4 wn” Zz ”) . = ” a = w”

RIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS SJIYVYUSIT LIBRARIES Sh

q ” > ” > axe wn =

ee <O a WO = ” -

oc . = oc = a — ~.

By = ile < < - ay

ac sa oc Ee ac = oe.

SNI_NVINOSHLINS S3IYVYGIT LIBRARIES SMITHSONIAN _INSTITUTION NOILNLILSNI_N\

Ai FA FS cs = aE = Je

ras = Oo Z a Oo es

w ss w — Y-4 w — ow

a 5 = = y,: = =

3 ee — > rad Ll: tk >

pe: - a — MV, we - 2

AG one Ww ae as

RIES SMITHSONIAN INSTITUTION NOILMLILSNI_ NVINOSHLIWS, S31Yvu a! LIBRARIES Sh

b = a z < : ‘ <

. Zz \ = = —

NS = MWS S 6 WA og |

wz E WY 2 =. SWrsos

z = * S Pa =

JLSNI_ NVINOSHLINS S3IYVYGIT LIBRARIES SMITHSONIAN INSTITUTION -NOILOLILSNI_N)

LIBRARIES SMITHSONIAN

LIBRARIES . SMITHSONIAN

= ” s z 5

~ WwW i

O 3 - 4 =“ =

fo] ‘ * 4 = =

ee S O 2

1 ae a aol Zz 2 +

RIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYvVYdIT LIBRARIES SI

f = Se ee = & 5

n- 9 ve fe) Oo = o

as a = = x =

ay ies, = so

NEeZGH = E : x a

sO ae “yy == a F m aD |

ae o z 1 o =

LILSNI_ NVINOSHLINS S31YVUdIT SMITHSONIAN INSTITUTION, NOILNLILSNI_ N

LIBRARIES

ITHSONIAN

YINOSHLINS S31uVvusl

MY INOSHLIWS

AITHSONIAN

VINOSHLIWS

MITHSONIAN